KNC report

KNIGHT NEWS CHALLENGE

A look at what we’ve learnedA review of the 2010 and 2011 winners

Commissioned by The John S. and James L. Knight Foundation • Prepared by Kenneth Dautrich, The Stats Group

Knight News Challenge Findings Report1

This work is licensed under the Creative Commons Attribution-NonCommercial 3.0 United States License.To view a copy of this license, visit http://creativecommons.org/licenses/by-nc/3.0/us/ or send a letter to Creative Commons, 444 Castro Street, Suite 900, Mountain View, California, 94041, USA.

ABOUT THE JOHN S. AND JAMES L. KNIGHT FOUNDATIONThe John S. and James L. Knight Foundation advances journalism in the digital age and invests in the vitality of communities where the Knight brothers owned newspapers. The Knight Foundation focuses on projects that promote informed and engaged communities and lead to transformational change.

Table of contents

Executive summary 3

2011 News Challenge Winners 6

2010 News Challenge Winners 8

Lessons Learned 10

Lesson one 11

Lesson two 13

Lesson three 14

Lesson four 16

Lesson five 17

Lesson six 18

Lesson seven 20

Lesson eight 22

2010 KNC Winner Profiles 23

2010 KNC Winner Profiles 62

Knight News Challenge Findings Report 2

Executive Summary

Disruption and innovation

have become regular features

of the news and media

landscape. Social media feeds and

newsreaders are replacing printed

words and pages. Ordinary citizens

with smartphones and Twitter or

Instagram accounts increasingly

stand in for trained reporters. Hacker

journalists—wearing the hats of

both journalist and coder—crunch

massive data sets to find the insights

buried within, as major news media

organizations struggle simply to keep

up with the crowdsourced pace of

social media.

That’s where the Knight News Challenge comes in. Launched in September 2006 by the John

S. and James L. Knight Foundation, the News Challenge invests in people who are testing

new ideas for engaging citizens with news and information. It is an open contest designed to

accelerate innovation in the ways that we create, consume, and share news and information

by developing new ideas to reach more people more effectively. In each round of the News

Challenge, Knight Foundation trustees approve the winners as recommended by Knight

staff, with the advice of outside advisers. Since its inception, the Knight News Challenge has

provided more than $37 million to fund 111 projects in the United States and around the world.


KNC AT A

GLANCE

5YEARS

27MILLION

DOLLARS IN FUNDING

76PROJECTS

SERVED


In 2010 and 2011, the Knight News Challenge supported a diverse set of

media innovations—from a platform to help local newsrooms use and analyze

municipal data to a tool to help journalists make sense of vast amounts of

social media activity. In Vermont, 2010 News Challenge Winner Front Porch

Forum uses an online platform to strengthen the sense of offline community

in towns and cities across the state. When Hurricane Irene produced record

flooding in 2011, Vermonters used the platform to organize community

response and to connect towns in need with volunteer help. Across the world,

in Indonesia, palm oil farmers use FrontlineSMS—a 2011 News Challenge

winner that uses mobile technology to share and disseminate community

information—to organize collective efforts to challenge encroachments on their

rights by big palm oil corporations.

Knight Foundation hired evaluation firm Arabella Advisors to explore the

innovations and impact of these winners. Arabella reviewed grant materials,

analyzed Web metrics and social media data, surveyed the winners, and

interviewed both winners and key informants in the field. Through that

research Knight discerned lessons about what contributes to a successful

media innovation. These include:

• Measure success based on how funding improves the field, not just on

the adoption or impact of individual projects: Innovators and their sponsors

often view wide-scale adoption and sustainable organizations as critical

measures of success, but these are not always the best barometers. Building

the capacity of innovators as leaders in their fields and strengthening their

networks of supporters and collaborators can be just as important.

• Target users with “a need you can feel”: Projects that have scaled based

their innovation on a core audience and proven need. However, a large

number of projects faltered because they developed a tool without first

identifying target users. Unless a media innovation addresses a proven need,

news organizations often cannot spend money and time on projects or invest

in the technical capacity to take full advantage of new tools.

• Be open to the idea that your project may appeal to a different audience

than you imagined: Some projects designed to help the media analyze and

visualize data struggled to find journalists and news organizations that would

pay for the products. Instead, the products have gained traction among clients


in other industries. Small budgets in journalism and a lack of technical understanding

among journalists can inhibit adoption.

• Spend the time to get the user interface right: An intuitive user interface is vital for

attracting and retaining users. But a simple interface can mask a high degree of planning

and technical complexity. Innovators should not underestimate the time and expense

behind developing such deceptively simple interfaces.

• Provide substantial support to grantees beyond money, such as creating a cohort

of peers and providing access to influential networks: News Challenge winners

expressed a desire for support that comes from access to advisers who operate within

the foundation’s network and a desire to share their experiences with other winners

through in-person convenings that encourage the development of new connections.

• Anticipate resistance to innovation and the disruption it will cause, and plan

around it: Innovations frequently shake up their fields and meet with entrenched

institutional resistance. Successful innovators anticipate such resistance and plan

accordingly.

• Identify the elements of a project that require full-time staff and those that

can be entrusted to volunteers—and invest resources accordingly: An active

community of users and evangelists can perform certain functions that are critical for

the development of a media innovation. Other functions can only be performed by

dedicated, compensated, full-time staff. Innovators should identify who can accomplish

which elements early in their project, and invest accordingly.

• Recognize the benefits and challenges of open source code: The News Challenge

requires winners to use open source code and to publicly release it. This approach

encourages iteration and improvement, but the benefits may be to the wider community

instead of the challenge winner, who may bear the cost of development.

The Knight News Challenge has evolved significantly since its inception. Knight

continues to review the challenge and learn from the winners to help news and

information industries navigate the disruption in traditional strategies and uncover new

models of sustainability.

In the pages that follow we provide additional detail on these lessons, ideas and

insights—as well as on the progress of each of the winners of the Knight News Challenge

from 2010 and 2011.


2011 Knight News Challenge winners

Project Grantee Innovation Current

Status

Grant

Awesome

Foundation News

Taskforce

The Institute on

Higher Awesome

Studies Inc.

A vehicle for issuing micro-

grants to support innovative

local journalism and civic

media projects

Active $244,000

DocumentCloud

Reader Annotations

Investigative

Reporters and

Editors (IRE)

A new DocumentCloud

feature designed to engage

readers by allowing them to

add notes and comments to

original source documents

Active $320,000

FrontlineSMS Social Impact

Lab Foundation

(formerly The

Kiwanja Foundation)

A platform that enables

journalists to more effectively

use text messaging to inform

and engage rural communities

Active $250,000

iWitness Adaptive Path A Web-based tool for

aggregating and cross-

referencing news events with

user-generated content

Closed $360,000

NextDrop NextDrop An interactive voice response

and text message-based

service that notifies residents

of Hubli-Dharwad, India, when

their water is available

Active $375,000

OpenBlock Rural University of North

Carolina, Chapel Hill

A standard process and

structure for scraping public

records that allows rural

newspapers to gather, format

and publish municipal data

through the OpenBlock

platform

Closed $275,000

Overview The Associated

Press

An open source tool that can

make patterns within large

document sets visible, helping

journalists find stories in large

amounts of data

Active $475,000


Project Grantee Innovation Current Status

Grant

PANDA Investigative

Reporters and

Editors (IRE)

A set of open source, Web-

based tools that make it easier

for journalists to clean and

analyze data

Active $150,000

Poderopedia Miguel Paz A crowdsourced database that

visualizes relationships among

the political, civic and business

elite in Chile

Active $200,000

The Public

Laboratory

The Public

Laboratory for Open

Technology and

Science

An online community and set

of toolkits that enables citizens

to gather environmental data

about their communities

Active $500,000

ScraperWiki ScraperWiki New journalist-specific

features within an existing tool

to collect, store and publish

data from across the Web

Active $280,000

Spending Stories Open Knowledge

Foundation

A tool for contextualizing

government spending data

and improving fiscal literacy

among journalists and the

public

Active $250,000

The State Decoded The Miller Center

Foundation

A digital platform for parsing

and displaying state codes,

making laws readable and

accessible to the average

citizen

Active $165,000

StoriesFrom The Tiziano Project A storytelling platform for

combining user-generated

content with professional

sources

Closed $200,000

SwiftRiver Ushahidi An open source platform

that helps identify trends and

verify user-generated content

emerging from mobile phones

and social media

Active $250,000

Zeega Media and Place

Productions

A platform to empower

citizens and local news

organizations to create

multimedia stories about their

communities

Active $420,000

Total $4,714,000

2010 Knight News Challenge winners

Project Grantee Innovation Current

Status

Grant

Basetrack November Eleven An online journal and social media

resource center providing contin-

uous coverage of the entire de-

ployment of a U.S. Marine battalion

to southern Afghanistan

Active

$202,000

CityTracking Stamen Design LLC A Web service and open-source

tools to display public data in

easy-to-understand, highly visual

ways

Active

$412,000

Front Porch

Forum

Front Porch Forum

Inc.

A network of online neighborhood

forums in Vermont that allow users

to read and share posts with their

neighbors

Active

$220,000

Game-O-Matic Georgia Tech Re-

search Corp.

A free, easy-to-use tool that al-

lows journalists to build cartoon

arcade games based on their news

content

Active

$378,000

LocalWiki WikiSpot An easy-to-use, open-source

“wiki” platform tailored to the

needs of local communities

Active

$360,500

NowSpots Windy Citizen Open-source software allowing

“real-time” advertising that can be

updated at any time by local busi-

nesses using social media

Active

$257,500

OpenCourt Trustees of Boston

University

A pilot project to demonstrate

how digital technology can in-

crease public access to the courts

Active

$250,000

PRX Story Ex-

change

PRX Inc. A crowd-funding platform that

allows local public radio stations,

producers and listeners to find and

help fund stories

Closed

$75,000

SeedSpeak Arizona State Uni-

versity

An application with mobile, Web

and widget components that

provides citizens an easy way to

suggest community improvements

to local leaders, volunteer groups

and each other

Active

$93,600


SocMap Society Technolo-

gies Foundation

A map-based social network

where users can browse news and

engage in civic action through an

online local community map

Active

$265,000

Stroome Stroome An online video editing commu-

nity which allows users to upload

content and collaboratively edit

Active

$230,000

TileMill Development Seed A suite of open-source tools that

local media can use to make

custoim, embeddable hyperlocal

maps

Active $76,960

Total $2,820,560

Project Grantee Innovation Current Status

Grant


Lessons Learned

The winners of the 2010 and 2011 Knight News Challenges encompass a diverse range of approaches, audiences, geographies, goals and technologies. However, the

progress and challenges the winners faced illustrate common lessons which may apply to other innovators who seek to improve the ways communities produce, disseminate and consume news and information.

10

Measure success based on how funding improves the field, not just on the adoption or impact of individual projects.

The best barometer of success isn’t the outcome of individual projects but the effects projects may have on their sectors or industries. Funders should focus on building the capacity of innovators as leaders in their fields or strengthening their network of supporters and collaborators for long-term impact—regardless of the sustainability of particular projects.

For example, in developing The State Decoded, a 2011 winner, Waldo Jaquith hoped to build upon work in Virginia to make state laws more readable and accessible to citizens. The goal was to create a platform that could be adapted to state codes across the country. In doing so, Jaquith became a leader in the open government field. His success is attributable to several factors. An active community of users supports The State Decoded, and the platform has been adapted for use in a number of states and municipalities across the country. But Jaquith also set very clear goals for the project, and most importantly, he stuck with his original timeline. He outlined a clear beginning, middle and end for his involvement in The State Decoded, and eventually handed off its development to the community of open government activists and hackers. This has contributed to Jaquith’s leadership within that community. He continues to use his prominence to advocate for greater governmental transparency. As his involvement in The State Decoded was concluding, Jaquith launched—with Knight Foundation support—the

Funders should

focus on building

the capacity

of innovators

as leaders in

their fields or

strengthening

their network of

supporters and

collaborators

for long-term

impact—

regardless of the

sustainability

of particular

projects.


U.S. Open Data Institute, which replicates a British effort to encourage governments and businesses to adopt open data standards as a way to promote economic growth, innovation and social change, demonstrating his ongoing leadership in the open government field.

Investments in leadership sometimes pay off significantly even when products are not particularly successful or widely adopted. Brian Boyer developed PANDA as a set of Web-based tools that could serve as a newsroom’s data library. As conceived, PANDA would help journalists import, search, share and work collaboratively with large public data sets. Although PANDA has received praise for its technical sophistication and its usability, newsrooms have not adopted it as widely as hoped. The underwhelming adoption rate is partly attributable to the fact that Boyer and his project team were not able to dedicate themselves full time to developing and marketing PANDA. However, as he developed PANDA, Boyer’s stature in data journalism rose. Based on his work at The Chicago Tribune—and, presently, in his role as news applications editor at NPR—Boyer became a leader in the field, someone who could help bridge traditional journalism with the more technically sophisticated aspects of data analysis and visualization. Today, PANDA is no longer in active development, and by conventional measures, it failed the test of sustainability. But the project strengthened Boyer’s position as a leader and advocate in the field of data journalism—an outcome with potentially farther-reaching implications than that of a single tool, even if the tool had been widely adopted.

Although

PANDA has

received praise

for its technical

sophistication

and its usability,

newsrooms

have not

adopted it

as widely as

hoped. The

underwhelming

adoption

rate is partly

attributable to

the fact that

Boyer and his

project team

were not able

to dedicate

themselves

full time to

developing

and marketing

PANDA.


Target users with “a need you can feel.”

Many News Challenge winners develop innovative tools or approaches that target journalists, their employers and other media organizations, but selling innovations to news organizations is extremely difficult because they may lack the money and time to spend on innovative projects or the technical capacity to take full advantage of new tools. The innovation may also be entering a market guarded by institutions that may be resistant to change. Fundamentally, unless an innovation addresses a pressing need, journalists and news organizations will not adopt it. In fact, innovators need to anticipate resistance, and create development and marketing plans that address it. Innovators may need to diversify their user bases beyond journalists and news organizations to promote wider adoption and project sustainability.

In many cases, media organizations—especially in small or medium markets—lack not just the need for innovative tools, but also the resources and capacity to support ambitious technology development. One of the 2011 News Challenge winners, Zeega, aimed to build a platform that enabled local news organizations to create multimedia stories about their own communities. By developing an easy way to combine video clips, audio clips and images from a variety of sources, Zeega would make it easier for news organizations to tell stories in different and compelling ways. Initially, the project team provided consulting services to local media organizations to help them produce customized multimedia experiences with the Zeega tool. But they quickly

13

found that providing custom consulting drained limited staff time and resources and detracted from their ability to develop Zeega as a product that could have appeal to a general audience. The local news organizations that Zeega had identified as its target users were not willing to pay for the tool. Zeega ultimately changed both its product and its business model. Zeega’s leaders now view the target audience as the wider tech-savvy population equipped with smartphones and tablets.

In other cases, a real need for a new tool might exist, but the barriers to its adoption might simply outweigh that need. This is especially true in data-driven journalism. ScraperWiki, for example, a 2011 News Challenge winner, received funding to adapt its tool to help journalists collect, store and publish data from across the Internet. But the project team found that news organizations were either unwilling to pay for the tool or that the learning curve was too steep. ScraperWiki has since developed a more user-friendly version of its tool, but adoption rates among journalists remain below expectations, and ScraperWiki is still dependent upon non-media corporate customers to support development costs.

Be open to the idea that your project may appeal to a different audience than you imagined.

In some cases, a project’s ultimate audience or user base can differ dramatically from that for which it was originally conceived or designed. Several 2010 News Challenge winners made significant changes to help their projects

In other cases,

a real need for a

new tool might

exist, but the

barriers to its

adoption might

simply outweigh

that need. This

is especially true

in data-driven

journalism.


gain traction. While developing their respective tools, the project teams behind Stroome and Game-O-Matic tried to broaden their original audiences from journalists and editors to include citizen journalists and casual users. CityTracking moved in an opposite direction: Finding that journalists were too broad of an audience, it now focuses on serving the need of more technically proficient developers.

Overview, a tool to help journalists visualize patterns within large sets of documents, also faced a choice about whether to continue serving its intended audience or to shift to a new model. However, the project leaders also had to weigh their own values about what they hoped to achieve within their own innovation, even if those values might steer them away from models that made more financial sense. From the outset, Overview’s target audience was journalists, and its mission was to empower them to tell stories that might otherwise remain hidden in large, inaccessible or disorganized document sets. As the tool was being developed, Overview received an increasing amount of interest from potential customers in finance, business consulting and the legal profession. Pursuing these clients, however, would have required a shift of emphasis, a shift of resources, and a shift in organizational structure. The project team considered reincorporating Overview as a for-profit venture, but they kept coming back to the same conclusion: Although they might be able to develop a for-profit venture to attract funding to finance additional development costs, this would necessitate a shift away from their original target users—journalists. The Overview team determined that they didn’t want to become “just another startup.” They wanted to focus on their original social-driven mission and their original users.

As the tool was

being developed,

Overview

received an

increasing

amount of

interest from

potential

customers in

finance, business

consulting

and the legal

profession.

Pursuing these

clients, however,

would have

required a shift of

emphasis, a shift

of resources,


Spend the time to get the user interface right.

User interface can play a major role in determining whether a media innovation is actually adopted by its audience—an interface that’s fun to use or saves the user’s time can make the difference between a tool that’s used and one that gathers dust. Among the innovations developed by News Challenge winners, the most effective interfaces frequently have been those that appear simple or straightforward. But such user-facing simplicity is hard to build. The user interface of Front Porch Forum, for example, was deliberately designed to be clean and straightforward, unadorned with extraneous features. Although it is an online tool, Front Porch Forum’s end goal is to strengthen the sense of offline community in Vermont towns and cities. The project team has designed the site’s features and functionality around this social formula by keeping the interface deliberately sparse. This allows users to get what they need from the site and build their offline community, while discouraging them from spending “all day in front of a computer.”

If media innovators aspire for wide adoption of their tools, they cannot overlook the development of an effective user interface; it’s often more important than the features or functionality of the tool itself. Indeed, according to Ian Bogost of Game-O-Matic, developing features and functionality may represent 80 to 90 percent of the effort in developing an innovative media tool. But that last 10 to 20 percent entails developing usability and polish, and that’s often the hardest part of bringing a tool to market. Given the fast pace of innovation in the media marketplace, News Challenge winners may only have one opportunity to release their tool for wide use.


Provide substantial support to grantees beyond money, such as creating a cohort of peers and providing access to influential networks.

Many 2011 News Challenge winners expressed a desire for a greater degree of support in building strong and resilient project teams with the skills necessary to develop and scale their innovations; in developing effective marketing strategies to find new users; and in planning for sustainability beyond the period of the News Challenge grant. While Knight may be capable of providing some of this support, access to its networks of thought leaders and advisers can be invaluable for grantees negotiating these issues.

Just as important to the News Challenge winners, however, was the expertise of other winners. The 2011 winners reported that the opportunities to interact directly with fellow News Challenge winners—such as events held in Cambridge, Mass., Palo Alto, Calif., and Miami—proved to be extremely valuable, especially sharing information with projects that were either in different stages or had experienced similar challenges. Several winners expressed the desire for additional opportunities to interact with, and learn from, their fellow News Challenge winners. The opportunities that were most valuable were the in-person events in which winners could build connections with one another, and discover new connections with winners working in seemingly different arenas.

The value of these in-person convenings of News Challenge winners extends beyond individual cohorts. The 2011 winners valued their interactions with News Challenge winners from other years, and would have welcomed greater opportunities to nurture those relationships. Winners said they were more likely to seek support and advice from


other winners via e-mail and other means if they first met in person and developed some degree of familiarity.

Anticipate resistance to innovation and the disruption it will cause; plan around it.

Whether it takes the form of a new product or tool to empower citizen journalists or a new process to engage consumers of news and information, a media innovation often enters a space that is already occupied by time-tested methods and approaches, and one that often is guarded by institutions that may be resistant to change. These institutions may not react kindly to new innovations invading their space, because the innovation disrupts their normal course of operations. Innovators need to anticipate this resistance, and create development and marketing plans that reckon with it.

The 2010 News Challenge winners were no strangers to resistance. OpenCourt, for example, sought to change the way that citizens of Massachusetts were connected to their judicial system by live-streaming court proceedings

These institutions

may not react

kindly to new

innovations

invading their

space, because

the innovation

disrupts their

normal course

of operations.

Innovators need

to anticipate

this resistance,

and create

development and

marketing plans

that reckon with

it.


and trials in Quincy. But this represented a fairly radical change in how the court system in Quincy interacted with the media and with citizens at large, and OpenCourt faced numerous lawsuits that attempted to prevent it from streaming trial footage. Ultimately, OpenCourt prevailed on appeal to the Massachusetts Supreme Court, setting the precedent that OpenCourt—or other innovators in Massachusetts—could install cameras in courtrooms and broadcast their proceedings on the Internet. It succeeded in part because John Davidow, the project director, anticipated the strong institutional resistance he would face, prepared for it and had the support to persevere in the face of litigation and delays. Perhaps most importantly, the project had the benefit of an established home—Boston University—which paid for OpenCourt’s legal expenses as it fended off resistance.

Basetrack represents another example of a News Challenge project that sought to shake up institutional norms. In its effort to create an online, social media reporting network, it embedded a team of reporters and photojournalists with the 1st Battalion, 8th Marine Regiment during its deployment to Afghanistan. The military has strict rules governing how journalists can embed with deployed units in combat zones, and it was no small achievement that the project was able to embed with the Marine unit in the first place. Only a few months after deployment, however, the Marines asked Basetrack to cease its project, due principally to concerns that the project’s location-based reporting was revealing sensitive information about the position of U.S. forces. If the Marines were uncomfortable with the location data that Basetrack was providing, however, they could have worked with the project to remove the potentially dangerous information. But fundamentally, the military was extremely wary about

Certain important

elements of a

project—such

as product

promotion

and content

creation—can

be outsourced

in some cases

to users,

evangelists, and

the open source

community. But

other critical

elements—such

as core software

development,

business

development,

and fundraising—

should generally

be entrusted to

dedicated, paid

project staff.



unconventional efforts to report news from the battlefields and there was a limit to how far Basetrack could push the military’s standard practices regarding journalists.

Identify the elements of a project that require full-time staff and those that can be entrusted to volunteers—and invest resources accordingly.

Many projects plan at the outset to rely upon a dedicated user community to refine and promote an innovation, and upon vocal evangelists to drive wider adoption of their tools. In many cases, user communities and evangelists can become indispensable (and inexpensive) cornerstones of a project, especially when a project is dependent upon open source development. But without a core group of paid staff with the skills, the time, and the incentive to devote themselves full time to a project, development of a tool can suffer. Certain important elements of a project—such as product promotion and content creation—can be outsourced in some cases to users, evangelists and the open source community. But other critical elements—such as core software development, business development and fundraising—should generally be entrusted to dedicated, paid project staff.


When it comes to staff, passion alone is not sufficient—full-time commitment is often necessary, along with the money to make that a reality. The Tiziano Project, for example, won 2011 News Challenge funding to develop and refine its proprietary storytelling platform into StoriesFrom, which would combine user-generated content with content from professional journalists to tell news stories in more compelling ways. Relying on the strong reputation of its existing platform and on the enthusiasm of the founders, the project team experienced initial success in terms of developing partnerships and launched its platform ahead of schedule. But it quickly faced challenges related to its staffing model. Prior to winning the News Challenge, the Tiziano Project team consisted of highly motivated volunteers. The team dedicated a portion of its News Challenge award to paying for a full-time project manager and to providing part-time compensation for other team members. But this ultimately proved to be a significant underinvestment. The part-time team members lost the sense of commitment and excitement they had possessed as pure volunteers, while not being compensated to a degree sufficient to capture their full attention and energy. In addition, the team did not invest in staff dedicated to fundraising or business development. They had assumed that once the initial partnerships were forged, users would find StoriesFrom, use the platform and organically raise the visibility of the platform. As it happened, without a full-time staff member dedicated to business development and partnership management, momentum behind the project quickly slowed. The initial enthusiasm that users and partners expressed for the project faded as well, and without the investment in full-time staff to carry the work forward, the project faltered.

It is entirely

conceivable that

the winner might

bear the cost of

developing open

source code,

without receiving

an equivalent

or offsetting

benefit, which

might accrue to

someone else

entirely.


Recognize the benefits and challenges of open source code.

The requirement that News Challenge winners use open source code and publicly release any code they create has definite advantages. It encourages iteration and improvement, and it can magnify the impact of the winners’ work. DocumentCloud—which first won the News Challenge in 2009—produced Backbone.js, an open source JavaScript library that has since become a fundamental and widely used component for building Web-based applications, and in the words of one key observer, has proven “sufficient to justify the entire cost of the News Challenge.” But the open source requirement is not an absolute good, especially for News Challenge projects that include the scaling of an existing product or tool and that already have an established method for code development and dissemination. Front Porch Forum, a 2010 winner, represented such a case, with the project team reporting that the open source requirement was a drain on valuable time and resources, and that it provided little—if any—value to the project.

It is also important to consider where the benefits of open source accrue. In some cases, the News Challenge winners themselves benefit from using and sharing open source code. In other cases, it is the wider community of developers that benefits most. It is entirely conceivable that the winner might bear the cost of developing open source code, without receiving an equivalent or offsetting benefit, which might accrue to someone else entirely. It is important to consider such implications on a winner-by-winner basis, and to be flexible with grant terms and conditions to create an arrangement that will be most supportive of innovators’ efforts. The open source requirement could also be improved and implemented in a way that grants more flexibility in the types of open source licenses that winners can use.

2011 KNC Winner Profiles


Awesome Foundation 24

DocumentCloud 25

Frontline SMS 28

iWitness 31

NextDrop 33

Open Block Rural 36

Overview 38

Panda 41

Poderopedia 43

The Public Laboratory 46

ScraperWiki 49

Spending Stories 52

StoriesFrom 55

Swift River 58

The State Decoded 60

The Knight News Challenge: A Review of the 2011 Winners 1

Awesome Foundation News Taskforce

PROJECT GRANTEE INNOVATION GRANT

Awesome Foundation News Taskforce

The Institute on Higher Awesome Studies, Inc.

A vehicle for issuing micro-‐grants to support innovative local journalism and civic media projects

$244,000

The Awesome Foundation establishes autonomous chapters of trustees in cities around the world that distribute monthly micro-‐grants to compelling projects in their communities. The foundation received Knight News Challenge funding to apply its model for community-‐based financing to the field of journalism and to open chapters with an exclusive focus on local news projects. THE INNOVATION

Each chapter of the Awesome Foundation awards one $1,000 micro-‐grant per month to an exciting local project or organization. Chapter trustees are given full autonomy over grant-‐making decisions, a structure which empowers them to use their local expertise to determine which projects would be most useful for their communities. The small scale of each grant also encourages effective and efficient projects that might be otherwise overlooked by larger foundations that typically give out larger grants. The Awesome Foundation has started two chapters dedicated to journalism innovation so far, in Detroit and New Orleans. Early micro-‐grants have been awarded to a wide range of media projects, including photo documentaries, print shops, and city guides. Ultimately, the Awesome Foundation aims to foster local news communities by scaling its News Taskforce model to more cities around the United States. IMPLEMENTATION

The first News Taskforce chapter was established in Detroit in January 2012 and awarded its first grant in March 2012 to the

Detroit Journal, for a short film series featuring everyday Detroit citizens. Because the Awesome Foundation is headquartered in Massachusetts, a Detroit-‐based staffer (referred to as “the Dean of Awesome”) was hired to oversee trustee recruitment and manage the logistical aspects of building a chapter from scratch. With only one journalist on the inaugural trustee team, the Detroit News Taskforce spent much of its first six months consulting with area journalists to set parameters around what would qualify as a journalism-‐related project for the purposes of their grant making. Ultimately, the trustees opted to broaden the scope of grant-‐eligible projects beyond newspaper-‐ and magazine-‐centric proposals to include any project focused


Awesome Foundation 24

DocumentCloud 25

Frontline SMS 28

iWitness 31

NextDrop 33

Open Block Rural 36

Overview 38

Panda 41

Poderopedia 43

The Public Laboratory 46

ScraperWiki 49

Spending Stories 52

StoriesFrom 55

Swift River 58

The State Decoded 60


on providing information to the Detroit community. Nevertheless, building relationships with local media organizations proved more challenging than originally anticipated. The recruitment of Detroit News Taskforce trustees represented a change from how earlier Awesome Foundation chapters had been founded. Typically, Awesome Foundation chapters form organically, when community members come together around a common idea or interest. In creating the Detroit News Taskforce, the Awesome Foundation team adopted a more top-‐down approach, actively recruiting people willing to serve as trustees for a chapter with a predetermined topic focus. As a result, trustee engagement and retention has been a particular challenge for the Detroit News Taskforce, with some trustees who were less engaged from the start leaving the organization once they realized how much effort they would need to put in to sustain the organization. The funding structure of the Detroit News Taskforce may have also contributed to that chapter’s difficulty with retaining trustees. At other Awesome Foundation chapters, trustees pay $100 per month to participate, and those trustee contributions make up the source of all micro-‐grant funds. However, the Awesome Foundation has used some of its Knight News Challenge funding to cover the full amount of the Detroit News Taskforce’s -‐-‐grants, so trustees aren’t required to make any contributions themselves. Although this strategy has removed financial barriers to trustee recruitment, it has also had the unintended effect of producing trustees who have been less invested in the organization over the long run. The Awesome Foundation is currently exploring new fundraising methods to ensure the long-‐term financial sustainability of its Detroit News Taskforce. One such method is the “Awesome Tax,” a form of crowd-‐funded investment in which the News Taskforce solicits contributions from non-‐trustee community members on a recurring monthly basis. Another challenge that the Awesome Foundation encountered was the degree of

hands-‐on support and engagement that the News Taskforce required. Typically, the Awesome Foundation applies a very decentralized model to its local chapters, with little direct engagement in local operations or funding decisions by the core Awesome Foundation team. But the News Taskforce in Detroit required a greater degree of support from the core Awesome Foundation team than they had anticipated. The chapter struggled with how to reconcile the foundation’s typical boundary-‐less model with the specific issue-‐area focus of the News Taskforce. As a result, the Awesome Foundation had to invest more time in providing hands-‐on support and clearer operating parameters for the News Taskforce. In January 2013, the Awesome Foundation created a second media-‐focused chapter in New Orleans. The foundation applied many of the lessons learned during the Detroit News Taskforce’s challenging first year to build a more optimal chapter structure from the outset. New Orleans chapter trustees contribute to the organization on a sliding scale, paying anywhere from $5 per month to $100 per month depending on financial ability. This trustee funding model has the benefit of nurturing ownership and responsibility among trustees while not limiting participation from less affluent members. In addition, in an effort to increase trustee retention and engagement, the New Orleans chapter funds civic media projects only eight months out of the year, leaving four months per year for trustees to award grants to projects that align with their personal passions but fall outside the realm of civic media. By allowing trustees to fund projects of personal interest for a portion of the year, their commitment to finding and funding civic media projects for the remainder of the year will be deepened and strengthened. REACH AND OUTCOMES

Despite its trustee engagement challenges, the Awesome Foundation has succeeded in establishing two active media-‐focused chapters with strong early patterns of grant making. Since its 2012 launch the Detroit News



Taskforce has awarded $24,000 in funding to 24 Detroit-‐area media projects. The Awesome New Orleans chapter has awarded $9,000 to nine projects since its first grant in April 2013, six of which have had a significant media element. Examples of funded projects include an initiative to raise awareness about the importance of voting in local elections, a newspaper supplement written by children and young adults, and a literacy and arts festival. The Detroit and New Orleans grant recipients have thus far met with mixed success. Some have won funding from other, larger foundations. Other project creators have had compelling ideas, but have ultimately lacked the necessary resources to scale their projects beyond the local level. Finding projects to support that are both relevant to a local community and have the business capacity to expand regionally or nationally has proved more difficult than the Awesome Foundation team initially anticipated. The Awesome Foundation team, however, has discovered that $1,000 awards have benefitted winners beyond the value of modest grant amount itself. The Awesome Foundation model has shown promise as a method of identifying innovators who are likely to do good work in the future, regardless of whether their winning project succeeds or fails. As its winner list grows, the foundation has also made a greater effort to connect winners with one another, and in some cases, this has resulted in winners sharing resources and offering mutual support. In 2014, the Awesome Foundation’s main goal for the News Taskforces is to sustain the Detroit and New Orleans chapters without grant funding with a combination of trustee contributions and local business sponsorship. The Awesome Foundation team also intends to be more deliberate about facilitating relationships between particularly promising winners and larger funders like the Knight Foundation.



DocumentCloud Reader Annotations


Document Cloud

Investigative Reporters and Editors (IRE)

A new DocumentCloud feature designed to engage readers by allowing them to add notes and comments to original source documents

$320,000

DocumentCloud won a Knight News Challenge grant in 2009 to build a tool that helps journalists analyze, annotate, and publish original source documents. The project has experienced a great deal of success: it continues to gain adoption in newsrooms and grow. DocumentCloud was also instrumental in the development of Backbone.js, which is one of the most important Javascript libraries used in web development today. In 2011, DocumentCloud again won the News Challenge, this time to incorporate the ability to add reader annotations to source documents—a new feature that would allow newsrooms to invite the public to annotate and comment on source documents. THE INNOVATION

DocumentCloud is an open source, web-‐based platform that helps journalists analyze, annotate, and publish original source documents. To date, almost 1,100 organizations use DocumentCloud to store and share source documents with readers. Journalists can already annotate documents using the tool, and many users have requested a similar feature that would allow them to add notes and comments to documents as well. Adding a reader annotations feature would allow DocumentCloud to be used not only to link stories to raw documents, but also to crowdsource document annotation, allowing journalists to review massive amounts of documents faster with help from the public. The feature will help journalists involve their readers in the process of reporting and analyzing news events and will improve DocumentCloud as a tool and resource for investigative reporting.

IMPLEMENTATION

DocumentCloud began through collaboration between journalists at The New York Times and ProPublica. Following the announcement of their second News Challenge award, the project changed hands when DocumentCloud was acquired by Investigative Reporters and Editors (IRE), a nonprofit membership organization based at the University of Missouri. As of early 2014, DocumentCloud has yet to deploy its public-‐facing reader annotations feature. It is still in the process of developing and beta testing the feature with several newsrooms. Several factors delayed its release, the most notable of which was the challenge of trying to accomplish three discrete tasks at the same time: maintaining the platform at its current level of functionality, managing growth of the user base, and adding new features and functionality. DocumentCloud struggled with whether to allow readers to comment anonymously and with determining the best way to integrate reader comments into news organizations’ content management systems. Like many commenting features, it decided to link readers’ comments to their social media accounts (Facebook and Twitter) so they could not remain anonymous. The project team created a test version of the annotations tool early in its two-‐year grant period and used journalists’ feedback to help shape further development. User feedback pointed out additional improvements and modifications needed to improve the functionality of both the public annotation tool and other elements of DocumentCloud. Feedback indicated that the team needed to rebuild its document viewer so that public annotations could be stacked in a



legible and uncluttered way, and that the project team needed to improve DocumentCloud’s mobile version to allow for easier viewing of reader comments. Meanwhile, DocumentCloud’s rapid growth placed additional demands on its technical system and on the capacity of its project team at IRE. With more users came technical challenges of needing to improve the platform’s speed and its capacity to hold larger, more complex document sets. DocumentCloud will be working on its sustainability planning with help from outside consultants throughout 2014 and is considering various models for generating revenue into the future. As of March 2014, it was still in the process of beta testing the reader annotations with partnering journalists and planned to release the feature later in 2014. Once released, IRE’s executive director, Mark Horvit, believes reader comments ideally will be project specific, and used in cases where news organizations would gather facts/analysis from readers or the readers’ opinions. For example, a newsroom may use the tool to allow readers to comment on the collection of Sarah Palin’s leaked emails, or to allow readers to flag items within public expenditure data.

REACH AND OUTCOMES

DocumentCloud underestimated the challenge of managing rapid growth while adding the functionality for reader annotations. As a result, it experienced major delays in launching a public-‐facing version of the feature. The project team believes the development and testing process has been beneficial overall as it helped to identify several modifications needed to improve the tool as a whole. DocumentCloud still plans to release a new version of its platform, complete with the reader annotations feature, in early-‐ to mid-‐2014.

The distinction between DocumentCloud as a project and the team’s effort to develop a reader annotations feature is important to keep in mind. The reader annotations feature is behind schedule and has not yet met expectations. But the same cannot be said for DocumentCloud as an overall platform. DocumentCloud is poised to become a standard tool for newsrooms around the world. By March 2014, DocumentCloud hosted more than 990,000 documents, comprising almost 13.5 million pages, for more than 1,000 organizations. The project’s website routinely receives over a million document views per week, with peaks of more than a million per day. With support from the Open Society Foundation, DocumentCloud is looking to scale globally, and is modifying the platform to work with additional languages.



Even with its success, planning for long-‐term sustainability is a key challenge for DocumentCloud. The project received a separate grant to fund its strategic planning work with a group of outside consultants.

Planning is still underway, and the team is considering options for generating revenue which might include the creation of paid add-‐on features or the creation of a paid platform targeted toward other industries.


FrontlineSMS


FrontlineSMS Social Impact Lab (formerly The Kiwanja Foundation)

A platform that enables journalists to more effectively use text messaging to inform and engage rural communities

$250,000

Mobile phones are increasingly common even in developing countries with low literacy rates and large rural populations. SMS and MMS messaging (text messaging) are similarly popular and are among the most effective ways to quickly reach large numbers of people in many communities. Although many tools for communicating with people via mobile phones exist, few SMS management systems are designed specifically for journalists and news organizations. FrontlineSMS was awarded News Challenge funding in 2011 to expand and improve its existing platform, which enables users in developing and rural areas to organize interactions with large numbers of people via SMS, and to tailor this platform to the needs of journalists and news organizations around the world. THE INNOVATION

Introduced in 2005, FrontlineSMS is an open source platform that enables users in areas with poor communications infrastructures to disseminate and exchange information with large numbers of people over cell phone networks without the need for the internet. The first version of FrontlineSMS was a free desktop application that allowed users to reach large groups via text messages, using just a laptop and a mobile phone. FrontlineSMS was awarded News Challenge funding to further develop its software for use by journalists and to work with community news organizations and radio stations to more effectively use text messaging to inform and engage rural communities. The project later developed FrontlineCloud, a similar, web-‐hosted platform that allows users

to log in wherever they have internet access and to run projects remotely.

IMPLEMENTATION

At the outset of its News Challenge grant, the FrontlineSMS team intended to expand its original application and release a specific plug-‐in for use by journalists and community news organizations. The team hired Trevor Knoblich as its media project director and revised its original plans to include a research and consultation phase to gather feedback on the needs of rural media outlets and organizations already familiar with FrontlineSMS. After surveys and extensive interviews with members of media outlets from around the world, FrontlineSMS found that news organizations hoped to use the tool in three ways: • To disseminate news headlines, tips, or

follow-‐ups to long-‐form pieces to large subscription lists

• To coordinate staff, freelancers, photographers, and citizen journalists

• To solicit requests for information via a dedicated phone line (“Text us if you see harassment in your neighborhood,” for example)

FrontlineSMS released the second version of its original modem-‐based platform in June 2012 using its News Challenge funding. Within the first fourteen months of its release, version two of FrontlineSMS was downloaded more than 150,000 times. The original FrontlineSMS tool used a modem that allowed a user to send only eight messages per minute. In speaking with journalists and other potential users about their



needs, however, the FrontlineSMS team realized that media outlets preferred an online mechanism for managing their mobile communications. News organizations also needed a tool that would allow them to send urgent news alerts to a larger audience more quickly. In response, the team began developing FrontlineCloud, the web-‐based version of FrontlineSMS. Unlike FrontlineSMS, FrontlineCloud requires internet access. But it provides news organizations and journalists with a more flexible option for disseminating news headlines and information. In early 2014, FrontlineCloud was still in the beta testing phase. The team was also working to build an interoperable product set that would allow users to smoothly transition between online use with FrontlineCloud and offline use with FrontlineSMS. FrontlineSMS continues to offer a range of premium user support and paid-‐for consulting services to provide an additional revenue stream to support its work. These services include mobile integration and program design assistance, staff training, software customization, dedicated technical support, and evaluation support. REACH AND OUTCOMES

As of January 2014, the second version of FrontlineSMS had been downloaded more than 177,850 times—more than seven times the number of downloads of version one. FrontlineSMS’s downloads continue to grow at

a steady pace, with an average of about 730 monthly downloads by journalists and others in the media. An estimated 14,500 journalists are using FrontlineSMS in 76 countries across the world, including Eastern, Central, and sub-‐Saharan Africa; Southeast Asia; Pakistan; Indonesia; the United States; and the Philippines. News organizations using the tool include rural radio stations in Uganda and Kenya; larger media outlets like the Kenya Star; and multinational news outlets such as the BBC, The Guardian, and Al Jazeera. In Indonesia, rural farmers, journalists from Internews, and environmental advocates are using FrontlineSMS to report, connect, and raise awareness of palm oil corporations’ destructive environmental practices. After one story by Ruai TV, the local palm oil company agreed to repair a road that had long been a source of contention with the community. Although the focus of FrontlineSMS’s News Challenge grant was to release a plug-‐in specifically tailored for journalists and community news organizations, the tool is actually used by both news organizations and the nonprofit community. Organizations working to combat malaria have used FrontlineSMS to connect people to health services in the Democratic Republic of the Congo. In December 2013, the project received a $1.5 million Google Impact Award for a three-‐year partnership with the nonprofit Landesa to help secure land rights for over 80,000 families

0

3000

6000

9000

12000

15000

FrontlineSMS Users

Esgmated Increase Media Users (Aggregate) Linear (Esgmated Increase Media Users (Aggregate))

Trend line (Aggregate Media Users) Aggregate Number of Media Users



in Odisha, India. FrontlineSMS also received awards from the Hewlett Foundation and the United Nations Democracy Fund to train civil society groups and governments in ways to use SMS to create more efficient service delivery mechanisms around the world. Looking ahead, the project team plans to add a missed-‐calls feature that provides users with a free and easy way to call organizations through their FrontlineCloud and FrontlineSMS accounts. The team continues to work on building an interoperable product set to allow for smooth transitions between FrontlineCloud and FrontlineSMS. And through the course of 2015, Social Impact Lab plans to support Frontline SMS in the process of forming its own independent organization, in the hope of attracting even greater investment in the platform.



iWitness


iWitness Adaptive Path A web-‐based tool for aggregating and cross-‐referencing news events with user-‐generated content

$360,000

For media outlets looking to supplement professional news coverage with citizen-‐generated content, finding relevant content can be a challenging task. Keyword searches and hashtags fail to differentiate between first-‐person accounts of a news event and secondhand observations. Some news organizations have built custom systems to collect crowdsourced media, but these tend to be cumbersome and resource intensive, resulting in little actual use. User experience firm Adaptive Path won News Challenge funding in 2011 to bridge the gap between traditional and citizen media through iWitness, a web-‐based tool that aggregates user-‐generated content from social media during big news events. THE INNOVATION

iWitness combined time indexing and geolocation technologies to allow users to search for citizen-‐generated content by both time and place. A date-‐time selector let the user search for events by hour and minute, and a map location box let users enter either a general city or a specific street address. When a major news event occurred (such as Hurricane Sandy hitting the East Coast in October 2012), iWitness could show users Flickr photos and Twitter messages posted from people at the scene, all aggregated into a single, easy-‐to-‐browse interface. Although an increasing number of services allow their content to be geotagged in this way, iWitness was unique in focusing on organizing data about news events. By showing the same scene from multiple social media vantage points, iWitness aimed to provide a new way for people to explore and experience the news. Its ultimate goal was to

make it easier for journalists to find and analyze meaningful citizen content about world events.

IMPLEMENTATION

Adaptive Path is primarily a design consultancy. Identifying a need for a different kind of expertise to develop iWitness, it partnered with New Context, a software development company, to carry out the technical work of building the iWitness tool. New Context developers recognized that in order for iWitness to be used by newsrooms, it needed to be something that non-‐tech-‐savvy journalists could easily manage. Additionally, staffing and funding constraints meant that once iWitness was released, opportunities to perform ongoing maintenance of a server-‐based tool would be limited. For these reasons, iWitness was built as an entirely browser-‐based application. The initial development process for iWitness was fairly smooth. The project timeline was extended four weeks beyond what had originally been planned—two weeks were dedicated to final technical iterations refining the finished product, and two more weeks were spent on marketing and promotion activities. The team worked with newsrooms at The



Washington Post, The Wall Street Journal, The Seattle Times, The Palm Beach Post, and the Sacramento Press to beta test the application. iWitness was released to the general public on June 12, 2012. A promotional video about the tool posted on Adaptive Path’s blog at that time received around 5,000 views. The visual design of the application didn’t lend itself to a mobile display, but it was supported on Google Chrome and Safari, and it was viewable on mobile devices such as the iPad. Unfortunately, iWitness hit a critical roadblock when Twitter changed its API in June 2013. The new version permitted only authenticated Twitter users to take advantage of the Twitter API; prior to that, using the Twitter API wasn’t dependent on a user signing in. As a result, the mechanism by which iWitness retrieved information from Twitter was essentially blocked. In its News Challenge application, the iWitness team acknowledged the risks that potential changes to the Twitter or Flickr APIs might represent, as well as the tool’s vulnerability to such changes. Unfortunately, when Twitter changed its API, the iWitness team lacked the funding to execute the extensive technical retooling of the application

needed to restore full functionality. Such retooling would have involved reengineering the product to support a server-‐based solution with ongoing maintenance and production demands. Consequently, the team decided not to overhaul its software to account for Twitter’s new API. As of March 2014, the iWitness tool has been fully disabled, and iwitness.adaptivepath.com returns a user to Adaptive Path’s website. REACH AND OUTCOMES

Before the change in the Twitter API undermined the tool’s technical underpinnings, iWitness was gaining notable traction. Within the first 11 months after its launch, the site received approximately 18,000 visits from 13,000 unique users. The professional organization Investigative Reporters and Editors reported that several of their members used iWitness to support their coverage of events such as the 2012 Newtown shootings and the 2013 Boston Marathon bombings. Currently, iWitness is non-‐operational, and team members have no plans to return to update the project.



NextDrop


NextDrop NextDrop An interactive voice response-‐ and text message-‐based service that notifies residents of Hubli-‐Dharwad, India when their water is available

$375,000

One million residents of Hubli-‐Dharwad, India, have water piped to their homes. Water is only available through those pipes for a few hours each week, however, and some residents must wait up to eight days between water deliveries. Each of India’s major cities faces similar water scarcity, affecting more than 100 million people. Project lead Anu Sridharan and her team created NextDrop to provide an immediate, accurate way for residents to know when water will be available. Leveraging the widespread adoption of mobile phones in India, NextDrop notifies residents when water will be available in their communities. THE INNOVATION

NextDrop is a platform that uses SMS messaging and interactive voice response (IVR) technology to notify residents of Hubli-‐Dharward, India just before their water becomes available. Prior to this service, residents were forced to waste hours each day waiting for water as printed newspaper notifications about water deliveries were often too outdated and inaccurate to be useful. NextDrop partners with the valve men who control a community’s infrequent flow of water and trains them to use the mobile-‐based platform to notify neighborhood residents via SMS when the water is turned on. NextDrop asks residents to respond, confirming that the water has arrived. The project received News Challenge funding to launch NextDrop’s work in Hubli and to develop the platform so that it might be customized and implemented elsewhere as a way of distributing other types of real-‐time community information.

IMPLEMENTATION

NextDrop launched in September 2011. It faced its first significant challenge when the Indian government passed regulations that same month restricting companies from sending bulk messages for commercial purposes between the hours of 9:00 p.m. and 9:00 a.m. The project team faced a decision: either stop sending SMS messages between 9:00 p.m. and 9:00 a.m., or gain an exemption from government authorities that would categorize NextDrop’s messages as “transactional,” rather than “commercial.” Typically, navigating the necessary bureaucratic hurdles to accomplish this would have taken months. NextDrop, however, had developed close partnerships with its SMS provider Netcore and the Hubli-‐Dharwad water utility. Together with these partners, NextDrop discovered that the new regulations did not apply to SMS messages sent by government agencies. And because the actual senders of NextDrop SMS messages were valve men employed by the Central Water Commission, this exemption could apply to NextDrop. Working with Netcore and its partners at the water authority, the team gained this exemption and returned to service after being shut down for only 12 days. The project’s success was similarly threatened by a sharp increase in SMS prices. The cost of sending a single text message increased five times in NextDrop’s first few years of operation, forcing the team to rethink its business model and find ways of cutting extra costs. NextDrop decided to halve its text messaging by sending only one message to users an hour before their water became available.



Despite these challenges, NextDrop was able to continue building upon its work in Hubli. One of the team’s key discoveries was that many customers preferred to use IVR technology as opposed to SMS messages. Although many of NextDrop’s customers could read SMS messages, many lacked the literacy skills to write an SMS to confirm the arrival of their water. More users were willing to pay for NextDrop’s IVR notifications than expected, and the project’s response rate among residents rose from 10 percent to 30 percent after introducing a “missed call” option. Through an external impact analysis, the project team also found that it was having the greatest impact on those who could not afford to pay others to collect their water while they themselves were away at work. As a result, NextDrop pivoted toward marketing to the working poor (and expanding its services to Bangalore, to better target this group) and moved to a freemium model, no longer charging customers for its most basic SMS water notification services. By early 2014, NextDrop has proven the value of its service, and it is in the process of strengthening its team’s capacity to build relationships with government officials and to brand and market the platform more widely. It is also in the process of becoming a paid, two-‐way platform for citizen-‐government communications. The project is working with Karnataka Water Supply and Sewage Board and the Bangalore Water Supply and Sewage Board to pilot the use of NextDrop in collecting feedback and reports of pipe damage and outages from NextDrop’s users. Although the project team is still early in the process of developing these services for utility companies, NextDrop believes its platform will prove replicable for other government services, and it is pursuing long-‐term contracts with water utilities as an ongoing source of revenue. The project team is also early in the process of exploring the possibility of marketing the NextDrop platform for politicians, who could use it to communicate with, and gain feedback from, their constituents.

REACH AND OUTCOMES

NextDrop aimed to be in use by at least half of Hubli’s households—around 33,000—by the end of its two-‐year New Challenge award. It did not meet that target, but it has nonetheless shown strong signs of growth. By March 2014, about 17,300 households in Hubli had registered with the service. Since transitioning to a “freemium” model, the project expects to reach its target within in the near future. The Karnataka Water Supply and Sewage Board and Bangalore Water Supply and Sewage Board have both purchased the platform’s utility services, and NextDrop is in talks with the Hubli-‐Dharwad Municipal Corporation about eventually scaling water alerts service to every Hubli household. Other cities’ commissioners have also approached the project team, expressing an interest in replicating the NextDrop model for other government services such as power and sanitation. Despite encountering various technical difficulties which resulted in instances of late and intermittent water notifications, NextDrop has largely been successful in providing reliable notifications for water delivery. Its external impact assessment showed that when used correctly, NextDrop allowed users to avoid contaminated groundwater, assisted them with rationing and water planning, and provided them with additional free time and relatively greater water security. In addition to providing water notifications, NextDrop’s utility services stand poised to improve communication between citizens and the Indian government, and ultimately improve Hubli’s infrastructure for water access and distribution.



The project intends to sustain itself by engaging in long-‐term contracts with water utilities, using its platform to collect feedback and reports of pipe damage and outages from NextDrop’s users. NextDrop also received funding from its partnerships with the Social Capital Partnership, Unilever’s Young Entrepreneurs Awards, and the Global System for Mobile Association, an association of mobile operators and related companies. Ultimately, NextDrop expects to eventually serve all 1.2 million citizens in Hubli-‐Dharwad and to scale to the entirety of Bangalore. Project lead Anu Sridharan hopes to scale to the entire state of Karnataka, India by 2015, and to scale globally, to other regions without continuous access to water, by 2018.



OpenBlock Rural


OpenBlock Rural

University of North Carolina, Chapel Hill

A standard process and structure for scraping public records that allows rural newspapers to gather, format, and publish municipal data through the OpenBlock platform

$275,000

OpenBlock is the open source software of EveryBlock, which won the 2007 Knight News Challenge. EveryBlock was an online platform that provided citizens access to hyper-‐local news and public data. As rural news organizations often lack the staff to make public data available and digestible, Ryan Thornburg of UNC Chapel Hill received 2011 News Challenge funding to tailor OpenBlock to the specific needs of rural communities and to develop a blueprint for deploying OpenBlock in rural newspapers across the country. THE INNOVATION

Prior to OpenBlock Rural, few tools or services existed to help smaller, rural news organizations efficiently gather, format, and publish public records on their sites. OpenBlock Rural aimed to increase rural communities’ access to local information and to strengthen their newspapers’ technical expertise by providing a tool that would allow them to collect, aggregate, and publish public data such as crime and real estate reports, restaurant inspections, and school ratings. The project also aimed to provide rural newspapers with a new way to generate revenue by allowing local businesses to sponsor data categories within the OpenBlock platform. IMPLEMENTATION

OpenBlock Rural set out to standardize the process and structure of scraping public records in rural communities, allowing these communities to then publish this data through the OpenBlock platform. The project team intentionally focused on a single, smaller partner—The News Reporter in Whiteville,

North Carolina—as it built and deployed its prototype. The project’s launch was delayed by several factors, the greatest of which was the difficulty it faced in acquiring digital public records. Rural communities often lack digital public records that are online, complete, and in a standardized format. Even when available, municipal data often suffered from misspellings and factual errors, and changes to the location or format of records caused OpenBlock’s scrapers (online tools used to extract information from websites) to break. In response, they launched open-‐nc.org, an online catalog of digital public data generated from state and local governments, in November 2013. Open NC was released as a free and open source Django web app in effort to assist other communities in making their data sets accessible to the public. In addition, the project faced concerns from both public officials and newspaper staff that citizens’ privacy outweighed their interest in government transparency. Other challenges included technical difficulties with geocoding news in rural areas (often due to incomplete data from the US Census Bureau) and higher-‐than-‐expected costs for local newspapers to host the application. OpenBlock Rural’s first year focused on overcoming technical challenges, most of which it did successfully. Its second year focused on the challenge of finding ways to use the platform to build a sustainable revenue stream for The News Reporter and other rural newspapers. Due to the continuing lack of available public records, however, OpenBlock Rural has no immediate plans to launch.



REACH AND OUTCOMES

OpenBlock Rural has yet to launch with The News Reporter or any other rural newspapers as of early 2014. Facing the challenge of insufficient digital public data, the project team instead turned to the mission of making it easier to find, request, and share digital public data within the state of North Carolina by launching open-‐nc.org. By March 2014, open-‐nc.org featured about 125 open data sets, including local arrest reports, property data, GIS files, and restaurant inspections, and was visited by 1,065 unique visitors both inside North Carolina and around the world. By providing easy access to the state’s public data, Open NC aims to support the transparency of its state and local governments, to lower the cost of watchdog reporting, and to increase innovation and economic development. As noted above, until more of the state’s digital public records are available online, OpenBlock Rural has no immediate plans to launch.



Overview


Overview The Associated Press An open source tool that can make patterns within large document sets visible, helping journalists find stories in large amounts of data

$475,000

As demand for government data and transparency increases, the release of large document sets is becoming more common. Whether from government transparency initiatives, leaks, or freedom of information requests, journalists have an increasing need to discern information from large data sets. Jonathan Stray, project lead at the Associated Press, conceived of Overview as a tool to help journalists explore and find stories within large data and document sets. THE INNOVATION

Several existing tools allow users to search large document sets for names and key words. But Overview aimed to be the first such tool specifically tailored to journalists’ needs, allowing them to discover new stories that might not even have been the basis for their initial search. Overview helps journalists discover stories hidden within massive document sets by using natural language processing to produce semantic maps that display the relationships among people, places, dates, and concepts. For example, a reporter analyzing large sets of emails can use Overview to sort the documents by topic, automatically grouping messages into threads and threads into subjects. Starting from a huge collection of unorganized files, Overview can automatically group documents by type and remove duplicates. Overview’s interactive system allows the user to explore these visualizations in order to detect patterns and reveal stories that might not have emerged through human sifting alone. The tool provides a way for newsrooms to gain a detailed understanding of the content within a large, unstructured database, allowing journalists to surface more original stories in less time.

IMPLEMENTATION

Overview set out to become a go-‐to tool for newsrooms seeking to explore and find stories within large sets of documents. After the debut of its first prototype with journalists at the National Institute for Computer-‐Assisted Reporting (NICAR) conference in 2012, it became clear that most users were unable to install the software or were unable to upload document files into the system. The project team hired a designer and spent months creating a web-‐hosted version of the tool, overhauling the user interface, changing its clustering algorithm, and completely rebuilding its document list based on feedback from early users. By the summer of 2013, Overview had addressed many of its largest usability problems and turned its attention toward marketing. The project team presented Overview to journalists through webinars, conferences, and blogs, as well as through the NICAR mailing list. It also implemented new CRM software for providing customer support. Throughout the two years of its News Challenge grant, Overview received interest from professionals within the fields of finance, business consulting, and government IT. The team considered two main options for developing Overview into a for-‐profit venture: selling the tool for use in monitoring brand conversations over social media, or selling it for law firms’ use in document review. Though these options increased the likelihood of sustaining the project, Overview ultimately decided against them, reasoning that this would divert resources away from developing the tool for their core audience of journalists.



After choosing to stick with journalists as its target market, the Overview team continued to customize its platform to fit journalists’ specific needs. The project originally anticipated that journalists would use Overview to summarize massive document sets. Journalists used the tool for a host of other scenarios, however, including when they needed to look for something specific within the data set, needed to classify and tag every document, or needed to filter out irrelevant material. Overview eventually implemented features that allowed users to complete these tasks. Today, the project is exploring several possible avenues for sustainability, including consulting to news organizations (training and providing support as they use the tool), selling the software as a service, and source licensing. The team is current transitioning to a paid model, which will charge a monthly subscription after a

30 day free trial. Overview expects this to cover its server operating costs, but will continue to pursue grant funding opportunities to cover developers’ salaries and the work of extending Overview’s API. REACH AND OUTCOMES

Overview encountered a tension between continuing to develop its tool for journalists and pursuing other markets to increase the likelihood of sustainability. It chose to focus on developing its tool for its core users: journalists. But Overview has struggled with user acquisition and creating a sustainable business model. From a technical standpoint, the project has been successful in creating a web-‐based tool that helps journalists successfully map the relationships between names, topics, and concepts in large data sets. Overview has been less successful, however, in gaining wider

0

500

1000

1500

Overview Page Visits

Page Visits Linear (Page Visits)

0

100

200

300

400

500

Feb-‐13 Mar-‐13 Apr-‐13 May-‐13 Jun-‐13 Jul-‐13 Aug-‐13 Sep-‐13 Oct-‐13 Nov-‐13 Dec-‐13 Jan-‐14 Feb-‐14 Mar-‐14

Overview Viewed Document Sets

Viewed document sets Linear (Viewed document sets)

Trend line (Site Visits)

Trend line (Viewed Document Sets)

Page Visits

Page Visits



adoption and use by news organizations. The project team noted strong levels of user retention, but they feel that they have yet to hit a critical mass of users, as too many journalists are still unaware of the tool. As of August 2014, Overview had more than 2,500 registered users on the web-‐hosted version of the platform and 10 million uploaded documents. On average, 15-‐20 users are active on any given day. Overview estimates that about half of its users are journalists, and the other half are professionals from within the fields of law, finance, and academia. Overview has experienced a steady increase in visits to its website, with more than 1,500 visits in March 2014. Spikes in traffic often correlate with blog posts that generate active discussions and that are reposted on outside sites, such as PBS’s Idealab. Stray has been focusing more energy on producing high-‐quality blog posts about Overview and the state of data-‐driven journalism that can drive traffic to the site. Overview has also experienced a steady increase in the number of people who view

document sets on the site: in March 2014, nearly 500 people logged into Overview to view document sets. Perhaps the most important metrics of the project’s success, however, are the number and quality of stories being produced using the tool. Stray described at least a dozen investigative stories developed using Overview, including a Newsday story created using Overview, which received a 2014 Pulitzer finalist award for Public Service. Another story from the Tulsa World used Overview to investigate $4 million misspent by the Tulsa Police Department on faulty squad car computers, via 8,000 emails obtained through a Freedom of Information request. In another case, a reporter from WRAL News in Raleigh Durham, NC used the tool to analyze 4,500 printed pages of emails from various government departments to uncover the root cause of technical problems that delayed delivery of food stamps to nearly 70,000 North Carolina residents. Overview allowed the reporter to finish this analysis in an afternoon, saving him or her weeks of work.



PANDA


PANDA Investigative Reporters and Editors (IRE)

A set of open source, web-‐based tools that make it easier for journalists to clean and analyze data

$150,000

Brian Boyer won the Knight News Challenge to create PANDA, a set of web-‐based tools to make it easier for journalists to work with federal, state, and city data. Smaller news organizations often lack the staff, knowledge, and tools to handle complex data sets. PANDA sought to help newsrooms share and make better use of public data, enabling more data reporting and stronger journalism. THE INNOVATION

PANDA serves as a newsroom’s data library, making it easier for journalists to import, search, share, and work collaboratively with large public data sets. The application also integrates data cleanup tools like Google Refine to help users find relationships among data sets and to help improve data sets for use by others. PANDA was designed to be used with Microsoft Excel, and to be easy enough to use to allow newsrooms without software developers to integrate it into their work. IMPLEMENTATION

While working as a reporter at the Chicago Tribune, Boyer and his project team needed to quickly search and share public data sets. The Tribune had its own tool for this, but it was difficult to maintain and reporters were required to update the site every time they found new data. The project team was awarded News Challenge funding to develop PANDA for The Chicago Tribune and for other newsrooms around the world. In an effort to understand and design the tool around their users’ needs, PANDA conducted extensive interviews with reporters and editors and distributed a survey through Twitter and the National Institute for Computer-‐Assisted

Reporting (NICAR) listserv. The survey focused on determining the technical aptitude of users’ newsrooms, the quantity of data they work with, and possible barriers to using the software. In February 2012, after six months of research and initial development, PANDA released a beta version of its platform. Among other features, PANDA allowed users to automate data imports, to search data sets using simple or complex search queries, and to set up automatic email alerts for news events related to newsrooms’ data sets. The project team aimed to market PANDA through social media, the NICAR listserv, and by conducting outreach and trainings at conferences. Because all four members of the project team held other full-‐time jobs while working on PANDA, turnover and time constraints were among the greatest hurdles to developing and marketing the tool. Around the time of PANDA’s release, Boyer left his position at the Tribune for a job as the news applications editor for National Public Radio. Developers Chris Groskopf and Joe Germuska also left the Tribune during the two-‐year grant period. In October 2013, PANDA revamped its website and marketing materials to target newsroom decision makers and to make a more focused case for data journalism, rather than concentrate its marketing efforts on data journalists themselves. Early users received it with excitement. However, PANDA continues to struggle to gain greater adoption in newsrooms. By early 2014, it has not received additional funding and is no longer in active development. Members of the original project team occasionally collaborate to fix bugs, and the open source community of PANDA users plans



to release translations of the software in additional languages within the coming months. REACH AND OUTCOMES

Without the dedicated time to promote the tool or a sufficient marketing budget, PANDA struggled to gain the level of newsroom adoption it had originally envisioned. Although the project team is unaware of the exact numbers of PANDA users, Boyer estimates that journalists from around two dozen newsrooms had downloaded the tool by late 2013, representing about a tenth of PANDA’s target adoption rate. At least four newsrooms are making heavy use of PANDA, including the Chicago Tribune, Tampa Bay Times, San Antonio Express News, and Dallas Morning News. San Antonio’s news team uses PANDA to store data sets such as public employees’ salaries and campaign finance reports. In one instance, San Antonio Express News reporters used PANDA to

quickly access state campaign finance records for a breaking news story, allowing them to produce a more detailed and time-‐sensitive piece than they would have been able to produce otherwise. Despite its slower uptake in the United States, PANDA has received interest from the international community, and the project team ultimately released versions in Spanish, German, Italian, and Portuguese. PANDA maintains an active online community through its Google group, though adding additional features or further developing the software would require additional investment. Independent of the original project team, one dedicated PANDA user from the Tampa Bay Times has sought funding to continue marketing the tool through videos and case studies that demonstrate its value.



Poderopedia


Poderopedia Miguel Paz A crowdsourced database to visualize relationships among the political, civic, and business elite in Chile

$200,000

In Chile, political and business relationships are often characterized by a lack of transparency. At the same time, the country has a journalistic culture that largely avoids reporting on (or even mentioning) the links among the country’s power elite. For the few journalists who have sought to report on powerful relationships in politics and business, gathering information about who is connected to whom has been a time-‐consuming and difficult endeavor. Poderopedia is a crowdsourced site that challenges Chile’s opaque power culture by mapping relationships among prominent Chilean leaders, making it easier for journalists to find and expose potential conflicts of interest. THE INNOVATION

The Poderopedia site consists of entries for people, businesses, and organizations. Each individual entry has a brief summary or biography, a tab listing connections, and a map of the person or entity’s notable relationships. It also contains links to relevant documents (such as a politician’s statement of assets) and the sources from which the entry information was taken. Poderopedia’s core staff writes many of the entries, but crowdsourced contributions are accepted as well, though all information they receive from outside parties is thoroughly fact-‐checked before being posted. Ultimately, the goal of Poderopedia is to transform the way money and power are discussed in Chile by setting an example of holding powerful people accountable.

IMPLEMENTATION

Initially, the Poderopedia team intended to launch the site within three to six months of receiving the Knight News Challenge grant. However, the team quickly realized it had underestimated how ambitious it would be to simultaneously build the technical aspects of a user-‐friendly website and develop all of the site’s initial content. This led to a lengthy extension of the project timeline. Poderopedia finally released its public beta in December of 2012, nearly one year after work on the site had begun. Due to the longer-‐than-‐expected development period, Poderopedia also spent significantly more of its Knight Foundation funding in its first year than it had planned. After Poderopedia’s release, the team’s second year focused on attracting users and building awareness of the site among journalists—a challenging task, given the project’s unforeseen financial resource constraints. When these constraints placed Poderopedia’s viability in jeopardy in early 2013, the team was forced to turn to funding sources other than the Knight Foundation, and secured $40,000 from Start-‐Up Chile, an accelerator program that aims to



0

10

20

30

40

50

Sep-‐13 Oct-‐13 Nov-‐13 Dec-‐13 Jan-‐14 Feb-‐14 Mar-‐14

Poderopedia Social Media Mentions

Other Facebook Twi\er

0

1000

2000

3000

4000

Dec-‐12 Feb-‐13 Apr-‐13 Jun-‐13 Aug-‐13 Oct-‐13 Dec-‐13 Feb-‐14

Poderopedia Site Visits

Site Visits Linear (Site Visits)

attract innovative, high-‐potential entrepreneurs to the country. Miguel Paz, Poderopedia’s founder, also began a fellowship with the International Center for Journalists and applied some of the stipend money to support the project. Although these strategies kept the project afloat, they also came at a cost, diverting some of the team’s energy from working on Poderopedia itself to grant reporting and fellowship project work. Poderopedia also faced a challenge specific to the media culture in Chile, where proper citation is not a central tenet of journalism and powerful media organizations often make efforts to block news stories that they don’t want reported. While several mainstream media organizations have picked up Poderopedia stories, they have rarely credited Poderopedia as a source. This has made it difficult for the Poderopedia brand to extend beyond in-‐the-‐know journalists to a general audience, even as Poderopedia’s founder has invested substantial time in marketing the site and it has done ultimately impactful work

driving transparency in Chile. As Poderopedia moved into its third year of operation, the project team had two primary goals. The first was to update the site’s open source code to make it easier for Poderopedia users to upload data in bulk and to model power relationships in other contexts outside of Chilean politics (such as other countries or industries). Updating the code would also make it easier for developers to modify and adapt the site for use in other countries, which had proven difficult thus far. The Poderopedia team had been aware of the need to update the site’s underlying code for some time, but lacked the funds to improve it. The team’s second goal was to reduce its financial dependence on grants, as applying for grant funding and approaching donors has continued to be a time-‐consuming challenge. Paz aimed to have no more than 50 percent of project income come from grants by the end of 2014 and was seeking funding through other revenue sources such as consulting to other organizations and by leading classes and trainings.

Trend line (Site Visits)



REACH AND OUTCOMES

Poderopedia’s site averaged about 570 visitors per day by March 2014. Notable traffic spikes tended to occur during major political events, such as a presidential election or a cabinet change, indicating that its mission to provide quickly accessible contextual information to reporters is being fulfilled on some level. Despite its readership, Poderopedia has struggled to attract contributors and is working to make changes to foster larger amounts of crowdsourcing. Poderopedia has received over 450 crowdsourced contributions from users, but the project team estimates that about 85 percent of its content was developed by the project’s team of editors. Beyond helping journalists cover large news events, Poderopedia has managed to break its own news stories. In one instance, Poderopedia first reported on a senator’s conflict of interest between his role leading a senate committee overseeing the logging industry and logging land he privately owned. Once Poderopedia published the story, another news outlet picked it up, which eventually led to a senate investigation. In a similar example, a congressman failed to disclose his shares in energy companies that would be affected by his energy committee, and Poderopedia exposed that link as well, which led to public outcry. This suggests that, little by little, Chilean political culture and the media that reports on it are becoming more transparent. Anecdotally, Poderopedia is a major though often silent contributor to that shift. Poderopedia has also done notable work expanding its mission to encourage media transparency, and not just in Chile. Paz established Poderopedia-‐branded workshops for teaching journalism throughout Latin America, in the process creating a community around the importance of transparency and using technology to tell stories. Paz also gained individual prominence as a champion of government and media transparency, with over 44,000 Twitter followers who receive his updates on Poderopedia and the broader aims

it supports. International interest in Poderopedia has grown, as well: a Venezuelan Poderopedia was scheduled to launch in April 2014, and groups in Spain, Colombia, Bolivia, and Puerto Rico were weighing the possibility of launching their own versions of the site.



The Public Laboratory


The Public Laboratory

The Public Laboratory for Open Technology and Science

An online community and set of toolkits that enables citizens to gather environmental data about their own communities

$500,000

The Public Laboratory (“Public Lab”) is a collaborative network that develops inexpensive and accessible do-‐it-‐yourself (DIY) scientific tools to engage citizens in solving local community challenges. Based on the success of its Grassroots Mapping project—in which volunteers mapped the Gulf Coast oil spill using helium-‐filled balloons and digital cameras—Public Lab won the Knight News Challenge to nurture and develop a community dedicated to expanding its set of DIY tools and promoting grassroots data gathering and research. THE INNOVATION

Public Lab designs and distributes kits of DIY tools with the goal of empowering citizens to investigate environmental issues in their local communities. Citizens without science backgrounds have previously lacked the money, skills, and technology to assess the environment around them, relying on government, corporate, and academic researchers to study and publicize matters of public and environmental health. Public Lab publishes instructions for building low-‐cost scientific instruments out of basic materials, enabling laypeople to collect crucial community data on their own. For a few hundred dollars or less, interested citizens can construct their own spectrometers for identifying oil contamination in water and soil, or near-‐infrared cameras for analyzing plant health. To reduce assembly time, they can also purchase kits for each tool from the Public Lab’s web store that include all the necessary materials. Public Lab provides detailed information on analyzing data from tools on its website, along with forums where community members can comment on tool

design and collaborate on potential improvements. IMPLEMENTATION

Public Lab launched with the goal of creating a community around DIY environmental exploration. Prior to receiving Knight News Challenge funding, the Public Lab team launched publiclab.org in January 2011 as the community’s online home. After winning the News Challenge, they planned to develop and post a new tool to the site every three months. Thanks in large part to development work that predated the News Challenge grant, by the end of 2011, the site already hosted instructions for assembling nine different scientific tools, including near-‐infrared cameras, balloon mapping kits, and hydrogen sulfide sensors.

While tool development proceeded on schedule during the early months of the grant, the seven founders’ geographical separation led to human resources complications—for instance, health



care and workers’ compensation were more costly given the distributed staffing model—and exacerbated the already steep challenge of aligning all of the founders’ visions for Public Lab. To address this, the Knight Foundation funded a facilitator for the team’s May 2012 in-‐person staff meeting, who helped the founders establish more effective strategies for collaborating at a distance.

Funding became a larger challenge for Public Lab in its second year. The founders had assumed that Public Lab would attract additional support beyond the News Challenge grant, but that support failed to materialize. The fundraising difficulties were largely due to a lack of staff capacity—the Public Lab team spent more time than they had anticipated in their first year building organizational infrastructure instead of nurturing long-‐term funding relationships. Recognizing the need for greater attention to funding, Public Lab hired a director of development, and initiated Kickstarter campaigns to sell and distribute retail kits of its scientific tools. The Kickstarter campaigns proved effective for introducing the project to technologically interested experimenters and early adopters, and succeeded far beyond staff expectations, with the balloon mapping kit and spectrometer kit combined generating over $150,000 in sales in 2012.

REACH AND OUTCOMES

Public Lab has built a substantial community around grassroots scientific tool-‐building and data collection. As of February 2014, publiclab.org had received over 540,000 unique visitors since the start of the Knight Foundation grant on September 1, 2011, and is averaging over 1,800 unique visitors a day in 2014. The more than 1,200 registered users of the site are notably active, having posted over 1,500 research notes and created over 600 wiki pages since the site’s launch. Public Lab has also garnered significant media attention, with mentions in The New York Times, The Boston Globe, Popular Mechanics, Wired.com, TechCrunch, Fast Company, and many other outlets.

The most deeply engaged community members are the Public Lab organizers, a group of 45 leaders who coordinate Public Lab activities in their local communities and tend to be the most frequent contributors to the site. Organizers have played an invaluable role in expanding the visibility of Public Lab in locations with no staff presence and in generating sign-‐ups for the organization’s 11 region-‐specific mailing lists. Public Lab’s team credited the geographic spread of its founders for this presence, and believed it allowed various regional groups to arise that would not have, otherwise.

0 1000 2000 3000 4000 5000 6000 7000 8000 9000

Sep-‐11 Dec-‐11 Mar-‐12 Jun-‐12 Sep-‐12 Dec-‐12 Mar-‐13 Jun-‐13 Sep-‐13 Dec-‐13 Mar-‐14

Public Lab Site Visits

Site Visits Linear (Site Visits)

Web links to Spectrometer Kickstarter

German website reprint of Technology Review article

Google announces map publication

Trend line (Site Visits) Site Visits



Enthusiasm within the Public Lab community has assisted staff in forming partnerships with organizations from the Gulf Restoration Network to the National Affordable Housing Network to the University of Massachusetts Amherst.

In accordance with its mission, Public Lab has strived to make its community as accessible as possible. Its site has a reputation as one of the friendliest open source communities on the web, and includes tool-‐building instructions in eight languages and counting. Offline, staff lead workshops to introduce the group’s data-‐collection methodologies to community members who don’t have internet access. Public Lab has also hosted three annual “Barnraisings”—conferences where participants gather to collaborate on new tools, research

directions, and projects. Fifty members attended the most recent November 2013 barn raising in New Orleans.

Public Lab was successful at fostering an engaged and involved community, but struggled with internal capacity issues early on in the grant, causing a backlog of new tools that the team was unable to launch. Despite this, Public Lab’s Kickstarter retail sales and Kits Initiative have helped to increase the size of their community, providing both a way to incubate projects and an additional source of funding for Public Lab’s newly approved 501(c)(3) nonprofit. The project has continued its growth well into 2014, hiring two new fulltime staff members and securing over $800,000 in funding, including another Knight News Challenge award in 2014 and federal funding from the Environmental Protection Agency (EPA).



ScraperWiki


ScraperWiki ScraperWiki New, journalist-‐specific features within an existing tool to collect, store, and publish data from across the web

$280,000

As more institutions make data available online, the potential to increase data-‐driven journalism grows immensely. Unfortunately, accessing and processing data is often tremendously difficult, particularly for journalists and citizens with limited programming skills. Few tools exist to help journalists find, access, and process obscure or complex data. ScraperWiki, a start-‐up in Liverpool, England, sought to develop journalist-‐specific features to facilitate the collection of information from across the web. THE INNOVATION

ScraperWiki allows users to collect, store, and publish public data, with a tool called a scraper. The data they scrape is then made available for others’ use. ScraperWiki was awarded News Challenge funding to add a data on demand feature that is specifically tailored for journalists, and to accelerate adoption of the platform in the United States by hosting journalism data camps. The new feature would allow journalists to request data sets and be notified of changes in data that might be newsworthy, and it would allow them to place data embargos that keep information private until after their stories break. ScraperWiki ultimately aims to allow journalists to produce richer stories and data visualizations by providing them with the means for accessing updated, aggregated public data. IMPLEMENTATION

The first goal of ScraperWiki’s News Challenge award was to add new, journalist-‐specific features to its existing platform. These features included a tool for on-‐demand, rapid-‐response data scraping; custom, private scrapers for a fee; assistance in developing public data sets; and a data alert service that notifies journalists

about changes in relevant data sets. ScraperWiki released its first iterations of these features in late 2011, and it used customer feedback and A/B testing to drive ongoing development.

The second goal of ScraperWiki’s News Challenge grant was to accelerate the adoption and use of ScraperWiki throughout the United States by hosting journalism data camps in New York, NY; St. Louis, MO; Washington, DC; San Mateo, TX; and San Francisco, CA. ScraperWiki kicked off its program of events in January 2012. To conduct these data camps, ScraperWiki partnered with newspapers and organizations such as The New York Times, the Chicago Herald Tribune, the Sunlight Foundation, ProPublica, Investigative Reporters and Editors, Spot.Us, the Centre for Investigative Journalism, Code for America, and the Online News Association. It also received support from journalism schools in its targeted states. During these events, journalists learned more about ScraperWiki and coding, and programmers learned more about ScraperWiki and data scraping.

A number of challenges impeded ScraperWiki’s ability to gain newsroom adoption. First, the project team found that newsrooms often had very little money and were unwilling to pay for data services. The team also encountered confusion about the platform’s free versus paid services. In addition, ScraperWiki found that its levels of newsroom adoption were low because its platform required a level of programming expertise and technical skill that few journalists and media professionals held. Often, journalists with coding experience also had their own internal tools for data mining and scraping. In response, the team decided to develop a new, more user-‐friendly platform that would address



these challenges and also allow more technical users to code within the browser application. It also introduced a community plan option, which allows users to get, clean, visualize, and analyze up to three data sets for free and to upgrade to a premium account if they agree to publish their data to Datahub.io, the Open Knowledge Foundation’s CKAN government catalogue. ScraperWiki released the beta version of its new platform in July 2013 and closed its original system. For journalists, the new ScraperWiki provides built-‐in data tools for specific tasks, such as scraping and searching for tweets, uploading spreadsheets, and summarizing large volumes of data through data visualizations. The new platform also allows more technical users to write their own code using GitHub, SSH, or the programming language of their choice. ScraperWiki continues to experience challenges in selling its platform and services to news organizations, where data would be used to support development of editorial content to create deeper, more compelling experiences for news consumers. It has experienced more success, however, in selling the tool to government agencies and corporate media clients. In this context, The Guardian, Channel 4 News, the UK Cabinet office, the United Nations Office for the Coordination of Humanitarian Affairs, Informa Agra, TraderMedia, and others have purchased subscriptions to the ScraperWiki platform. REACH AND OUTCOMES

ScraperWiki succeeded in building versions of its software for use by journalists and news

organizations. However, it has struggled to gain traction with journalists because few newsrooms are willing to pay for data services and because use of the platform previously required significant technical ability. The release of the new, more user-‐friendly system in July 2013 and the introduction of a free community plan subscription model were attempts to address this challenge and to make ScraperWiki more useful for investigative journalists. As of February 2014, only about 200 data journalists have registered for the tool. ScraperWiki as a whole, however, averages around 3,000 new registered users a month, many of whom are corporate media clients and government agencies. By March 2014, journalists from The Guardian, the Chicago Tribune, and the Texas Tribune had used ScraperWiki to produce stories. The Guardian used ScraperWiki in a front-‐page story revealing corporations and interest groups that channeled money to buy influence among UK’s parliament. Reporters used ScraperWiki to collect data located on different services from registers across parliament, and to aggregate it into one source table which could be viewed in an automatically updated spreadsheet or document. ScraperWiki has yet to achieve similar levels of traction within the United States, but it has succeeded in sparking connections and collaboration between over 500 journalists and developers through its data journalism camps. As the result of a connection made during ScraperWiki’s data journalism camp, two attendees Brian Ableson and Michael Keller have gone on to collaborate on various open news projects. The pair produced an interactive news app published in the Daily



Beast displaying the size and locations of mass-‐shootings within the United States.

Today, ScraperWiki continues to promote its platform to both journalists and corporate communications clients, relying on its consulting work and managed services in other

industries to help sustain the development of its tools for investigative journalism. ScraperWiki continues to market the services of its media tool pack to journalists, and it plans to conduct additional market testing throughout 2014.



Spending Stories


Spending Stories

Open Knowledge Foundation

A tool for contextualizing government spending data and improving fiscal literacy among journalists and the public

$250,000

News stories about government spending are commonplace, but without the appropriate context, it’s often difficult for readers to understand what the spending data means. The UK-‐based Open Knowledge Foundation, an international leader in organizing and interpreting open data, won Knight News Challenge funding to create the Spending Stories project. The goal of the project was to develop software applications to connect news stories to government spending information to provide quick access to much-‐needed context on public finance figures, thereby improving fiscal literacy for citizens and journalists alike. SpendingStories.org was to be the key application that resulted from this effort. THE INNOVATION

The ultimate objectives of Spending Stories were threefold: to give additional context to government spending numbers in the news; to make available more and higher-‐quality data about public spending; and to help people use that data once it becomes public. To achieve these objectives, the Open Knowledge Foundation introduced upgrades to its existing OpenSpending.org site, as requested by community users, and developed a free-‐standing Spending Stories application. In practice, Spending Stories puts public finance numbers in perspective by allowing users to compare spending figures cited in one news report with amounts reported elsewhere. For example, entering “1.0 million British pounds” into the search function in Spending Stories informs the user that that is the amount David Cameron has spent renovating Downing Street since his election. It is also five percent of the cost of the 2012 wedding of Prince William and

Kate Middleton. Users can visualize relationships between public spending figures on a scale or laid out in a card format with the most relevant stories appearing first. Users can also filter stories for those relevant to their interests, click to the original news story behind any number listed, and contribute stories to the database. IMPLEMENTATION

Although the goals for Spending Stories have remained constant, the format of the project has changed significantly over the course of the Knight Foundation grant. For the first year and a half of their grant, the Open Knowledge Foundation primarily focused on upgrading OpenSpending.org, the site which would serve as the hub for Spending Stories’ source data. Originally, the team planned to create a browser plug-‐in for journalists to embed in media sites that would systematically link spending figures to their source data. Doing so would have made Spending Stories one of the few media outlets linking public spending numbers back to their primary sources, rather than to other websites. For both technical and practical reasons, a browser plug-‐in to gather and source public spending data would not



have worked as well outside the United Kingdom, where such data is more readily available than it is in other nations. As a result, the project team decided to shift the site design to a human-‐generated matching system, in which users find news stories that mention spending figures and submit them for inclusion in the website. The revised design makes the site functional internationally while also creating a personalized filter so users aren’t overwhelmed by the number of stories on the site. The intended audience for Spending Stories has shifted as well. When it was conceived, Spending Stories was meant to be a resource for journalists. The project team expected that reporters would regularly visit the site to add context to their stories and conduct investigative work. With this in mind, the Spending Stories team budgeted to pay for a rotating team of journalists to contribute to a blog that would contain short videos and detailed commentary on spending in key issue areas. For several reasons, this model proved more difficult to execute than the team had anticipated. Most journalists had only minimal time to contribute to the project, and few were willing to invest the substantial training and effort required to work with Spending Stories’ complex datasets. Journalists also expressed greater interest in the browser plug-‐in idea. As such, the Spending Stories creators decided to focus on advocacy and non-‐governmental organizations whose interest in Spending

Stories stemmed from the fact that the app could be used to help with budgeting decisions in the developing world. Unlike journalists, these groups were also much more willing and able to partner with Spending Stories to explore and manage the project’s datasets. The changes to the project resulted in a significant alteration of the project timeline. After several internal reboots, including staff changes at the beginning of 2013, the Spending Stories application launched on November 21, 2013. Since then, efforts to attract users have involved promotion at in-‐person events and paying an outside contributor to populate spendingstories.org with an initial set of stories to generate interest. As of March 2014, the Spending Stories team was exploring other, less labor-‐intensive methods for adding stories, such as enabling any story tagged with #spendingstories on Twitter to be placed in a queue to be posted on the site, potentially by volunteer moderators. The team was also applying for additional funding outside of the Knight Foundation to support a set of targeted workshops on using Spending Stories and to evaluate the possibility of adding a leaderboard to the site to further incentivize user contributions. REACH AND OUTCOMES

In the first two weeks after spendingstories.org launched, the site averaged 69 visits a day. From December 2013 to March 2014, visitors

0

100

200

300

400

Nov-‐13 Dec-‐13 Jan-‐14 Feb-‐14 Mar-‐14

Spending Stories Site Visits

Visits Linear (Visits) Trend line (Visits)



dropped to about eight per day. Beyond basic tracking of site traffic and bug reports submitted by users, the Spending Stories team has had difficulty collecting information about its users, and team members acknowledge that this is an area where the project has room for improvement.

As of March 2014—only four months after launch—it was too early to judge the ultimate success of the Spending Stories site. Traffic to spendingstories.org has so been light. But visualizations created through OpenSpending.org—the hub for Spending Stories’ source data—have been embedded in prominent outlets including the Guardian, Le Monde, and Liberation, among others.



StoriesFrom


StoriesFrom The Tiziano Project A storytelling platform for combining user-‐generated content with professional sources

$200,000

Among the dozens of organizations conducting community journalism projects around the world, many are poorly funded or lack the leadership necessary to sustain a beneficial program. Based on the success of the Tiziano Project’s 360° Kurdistan, an immersive multimedia web platform for exploring the cultures of Iraqi Kurdistan, the group won the 2011 Knight News Challenge funding to redesign its proprietary 360° platform into StoriesFrom, a tool that community journalism programs in any local community could replicate. StoriesFrom was intended to give these programs a high-‐quality and convenient way to display compelling content while pairing the work of community and professional journalists.

THE INNOVATION

For 360° Kurdistan, the Tiziano Project team conducted a three-‐month citizen journalism training program in Iraq, matching trainees with professional journalist mentors and presenting their work on a single, interactive site. Their efforts on the project were recognized with numerous honors, including the Gracie Award for Outstanding News Website, the SXSW Interactive Award for Best Activism Website, the Community Collaboration Award from the Online News Association, and a Webby Awards honor for the Charitable Organizations/ Nonprofit category. StoriesFrom (initially titled the Tiziano Project 360°) proposed to build on the Tiziano Project’s prior work in several ways: it would redevelop the platform to make it replicable by organizations conducting similar workshops around the world, expand the platform to incorporate mobile technology, and create an

interactive world map to serve as a hub for projects developing StoriesFrom sites in their communities. The ultimate goal of the project was to enable local journalists to tell the stroies of their communities by improving the ways they could deliver news and information to larger audiences. In the process, they hoped the project would help shape public perceptions of regions that often receive one-‐sided coverage from Western media outlets. IMPLEMENTATION

The StoriesFrom team established relationships to pilot the project with several important organizations early in the grant period, among them the National Constitution Center, the National History Museum of Latvia, the Afghan Film Project, Machschava Tova, Media Art Xchanges, and the Fernando Pullum Community Arts Center. Pilot projects involved training students on reporting and media creation techniques, with the goal of empowering them to communicate stories of their communities online. In Latvia, 21 students learned how to report on local civic engagement. In Afghanistan, students were taught how to report on key community themes. On the San Carlos Indian Reservation in Arizona, five Tiziano Project mentors instructed 25 students in photography and multimedia creation skills. At the same time, StoriesFrom developers were constructing the web platform that formed the cornerstone of their proposal. The StoriesFrom beta site (www.storiesfrom.us) launched on July 7, 2012, two months ahead of the team’s original schedule. The timeline was accelerated after the team received an invitation to present StoriesFrom at the Dokufest International Film Festival in mid-‐July 2012. The new site was fully redesigned for HTML5 and optimized for the



iPad. Initial content consisted mainly of student work from the six pilot projects, along with the planned interactive map populated by the projects and curated tweets from around the world. Despite its early successes forming partnerships and meeting its launch goals, the StoriesFrom team faced challenges throughout the development process that only grew in magnitude after the site was launched. One large problem was with the structure of the team itself. Prior to winning the Knight News Challenge, Tiziano Project team members were all motivated volunteers. When the Knight Foundation funding came through, it was used to budget for one full-‐time project manager and part-‐time compensation for other team members, which had the unintended effect of demotivating formerly enthusiastic contributors while not providing enough of a financial incentive for them to fully commit to the project. Additionally, the part-‐time team model meant that crucial project roles, such as business development, were only being carried out half the time. The absence of a full-‐time business development employee took a significant toll on the long-‐term viability of StoriesFrom. The pilot partnerships the project had secured, while encouraging, were not designed to be maintained over a long period. The team had taken a “if you build it, they will come” philosophy toward their platform, assuming that the site needed to be launched before

promoting the site and reaching out to more potential partners. Instead, as the StoriesFrom team found itself in need of new partners, it became apparent that attracting interest in the site and additional funding would require significant work beyond release of the site itself—and that they lacked the necessary resources for outreach. As a result, work on StoriesFrom has stopped, with no current plans to revive the project. REACH AND OUTCOMES

The StoriesFrom site saw modest web traffic at the outset—1,600 unique visitors came to the site in July 2012, the month of its launch. Since that time, however, interest in the site has largely dropped off, with only a few traffic spikes around a few key events—for example, recognition for its Webby Award and a related Twitter mention brought 2,100 new visitors in March and April of 2013. Perhaps the biggest positive impact of StoriesFrom has come from the students it trained during the pilot phase. In follow-‐up surveys, pilot participants reported that they still use the skills they acquired during their StoriesFrom workshops, and that the skills have given them access to new job opportunities they wouldn’t have been qualified for otherwise. Their responses suggest that the training models employed by StoriesFrom staff were highly effective, even if the team was unable to sustain those trainings.

0

50

100

150

200

250

Jul-‐12 Sep-‐12 Nov-‐12 Jan-‐13 Mar-‐13 May-‐13 Jul-‐13 Sep-‐13 Nov-‐13 Jan-‐14 Mar-‐14

StoresFrom Site Visits

Visits Linear (Visits)

Webby Awards recognition and Twitter mention

Trend line (Visits)

Webby Awards recognition and Twitter mention

Visits



Instead of searching for a new sustainability model for StoriesFrom, which would likely require a $300,000 to $400,000 annual budget to reactivate, StoriesFrom creator Jon Vidar is applying his experiences by founding Uncharted Digital, a creative agency for storytelling

ventures. Uncharted Digital’s development team is based in Ukraine for cost effectiveness, and its American staff members are all full-‐time employees. The company is currently working on projects with the Tribeca Film Festival, the Sundance Film Festival, the Ford Foundation, and the Open Society Foundations.



SwiftRiver


SwiftRiver Ushahidi An open source platform that helps identify trends and verify user-‐generated content emerging from mobile phones and social media

$250,000

Ushahidi—a Knight News Challenge Winner in 2009—won the News Challenge again in 2011 to build on its past efforts to collect citizen-‐generated information originating from global crisis situations, such as the Kenyan election crisis in 2008 and the earthquakes in Haiti and Japan. As news events unfold, users of mobile phones flood the internet with firsthand accounts of these events. SwiftRiver aimed to help curate and verify this information by parsing it and evaluating its sources. THE INNOVATION

As the number of people who contribute newsworthy content grows exponentially with the growing use of mobile technology, the challenges facing journalists have shifted increasingly from problems of distribution to problems of discoverability and trust. Mobile phones and the social web allow citizens to report on major events and crisis situations around the world, but few tools exist to help journalists filter and differentiate this information based on accurate and trustworthy sources. Working across email, Twitter, web feeds, and text messages, SwiftRiver aimed to allow journalists, NGOs, government agencies, bloggers, and other organizations to identify trends and to evaluate information based on its creator’s reputation. IMPLEMENTATION

Ushahidi encountered a number of challenges in the development of SwiftRiver. Among the largest of these was building the technical infrastructure needed to store and sort massive amounts of data. Hiring engineers with the technical expertise to build systems of this scale required more funding than Ushahidi had anticipated. At the outset, Ushahidi hired a

small team of US-‐based engineers in Silicon Valley to develop the SwiftRiver platform, but it was unable to afford them as time went on. SwiftRiver’s next developers were based in Kenya and worked remotely, causing challenges as the teams worked across time zones. The project team also realized that the technology needed to create a simple tool to validate sources within massive streams of social media data may not be available or easily developed, given the team’s financial constraints. As such, the team’s conception of SwiftRiver evolved. Instead of focusing on creating a stand-‐alone platform that relied upon an automated system for verifying citizen-‐generated information, the SwiftRiver team focused on creating a set of tools to crowd-‐source the task of filtering citizen-‐generated information. These tools would exist within Ushahidi’s existing suite of services. The most popular of the six tools that ultimately comprised the SwiftRiver platform included its semantic tagging and geolocation tagging APIs. SwiftRiver’s semantic tagging application, Chambua, allows users to analyze text and extract words and terms that can be classified as people, places, and organizations. It can also recognize nationalities, religions, expressions of time, and monetary values. Chambua cannot fully and completely verify sources in a data stream, but it provides users with a tool that helps to sort and organize the data—which is a first step toward making sense of it. Ushahidi released SwiftRiver publically as an open source product available as part of its existing suite of tools in June 2013. Interest and demand for the tools remains high, and Ushahidi anticipates paying customers or the larger developer community will support and



further develop the platform. The code generated and lessons learned through its development will help inform version three of the Ushahidi platform, scheduled to be built in 2014. REACH AND OUTCOMES

SwiftRiver aimed to create an open source platform that filters information about major events and crisis situations, and identifies and verifies the most authoritative and accurate of these accounts. By doing so, it aimed to increase the viability of crowdsourced data collection as a methodology and practice for journalists. The project succeeded in building tools that help crowdsource the task of filtering large data sets, but it fell short of developing a tool that can filter and verify accurate sources within massive, real-‐time streams of data on its own. Through the development process, the SwiftRiver team realized found that it underestimated the technical challenges in creating a simple, mechanical tool for validating sources and making sense of massive data sets. These challenges proved to be insurmountable for Ushahidi, given its financial constraints. The team initially focused on marketing SwiftRiver to larger organizations, but soon scaled back to targeting smaller municipalities.

The largest deployment of SwiftRiver to date is in Pierce County, Washington. In November 2012—even prior to the public release of SwiftRiver as part of Ushahidi’s existing suite of tools—Ushahidi began collaborating with Pierce County to develop FirstToSee, an application that uses the SwiftRiver platform to provide emergency managers in the Puget Sound area with an efficient and effective way of responding to citizen-‐reported incidents on social media. FirstToSee will eventually be made available to other regions outside of Puget Sound. Pierce County also uses SwiftRiver to keep tabs on other relevant social and cultural events and conversations, positioning the county to better engage and respond to citizen voices in a transparent, proactive, and interactive manner. Although the demand for and interest in a more advanced tool that validates sources on its own remains high, Ushahidi lacks the funding and staff needed to further build, extend, and manage the project. SwiftRiver is currently part of Ushahidi’s business products toolkit, available for deployment by paying clients. Using revenue generated by SwiftRiver, Ushahidi plans to eventually develop the existing tool into a cloud-‐hosted platform.



The State Decoded


The State Decoded

The Miller Center A digital platform for parsing and displaying state codes, making laws readable and accessible to the average citizen

$165,000

Many state codes—the laws that govern every state—have been online for several years. But in most cases, they only exist in inaccessible formats. Many are only available as PDFs, making them difficult to search and interpret for even experienced legal professionals and journalists. The State Decoded aimed to create a standard platform, deployable in every state, to parse the text of a state’s code and build a website where ordinary citizens could quickly access contextualized information about the laws that govern them. The project was led by Waldo Jaquith, a programmer and open government advocate who had developed an early version of The State Decoded for Virginia. THE INNOVATION

The State Decoded makes laws digitally accessible by providing a set of tools to create individual websites for each state’s code. Prior to The State Decoded receiving Knight News Challenge funding, much of the work of the open government movement had been focused on clarifying the political process of lawmaking, as opposed to making existing laws more readable. By prototyping the creation of state code websites with basic search functionality, embedded legal definitions, and a rich API, The State Decoded represented one of the first major attempts to improve the user experience of citizens who are interested in accessing and understanding current laws. The ultimate goal of The State Decoded was to change people’s relationships with the laws that govern them by equipping average citizens with information about how their state works, how their government functions, and the laws that impact their daily lives. Achieving that objective would also carry significant benefits for

journalists, giving them a reliable source of information on current laws when researching articles and enhancing the quality of the information media outlets provide about the state code. IMPLEMENTATION

Development of The State Decoded has largely proceeded as its founder initially envisioned. Building off of pre-‐Knight News Challenge work, Jaquith and his team launched Virginia Decoded in March 2012. The next step was to rework the Virginia-‐specific code so it could be applied to other states. That work began with the GitHub release of Version 0.1 of The State Decoded’s source code in May 2012. Subsequent updates to the code were released every one to two months through the end of 2012. To Jaquith’s surprise, interest in the pre-‐release software was much higher than he had anticipated. Early contributors to the code on GitHub proved to be valuable development partners, providing a number of suggestions for improvement that were incorporated into later versions by the core team. To accommodate the volume of community feedback, the time between releases increased in 2013 as each code release grew more ambitious. Version 0.6, released in February 2013, established a public API for State Decoded sites and created a standard XTML format for importing laws. Changes in version 0.7 were more substantive than in all previous versions combined, and consisted of optimizations for speed, efficiency, and navigability. The State Decoded released Version 0.8 in November 2013. Jaquith intends for this to be the final release before it releases version 1.0.



Although the project has largely stuck with its original plan, a few factors have led to slight modifications. Several groups expressed an interest in seeing municipal codes given the same digital treatment as that of states, and later software releases supported that interest. Additionally, an intention to hire a domain expert in the typography of legal texts was scrapped when it became clear that only a few such experts exist, and that their time would have been prohibitively expensive. Instead, the team worked with a designer with experience in typography and absorbed the additional cost of that person learning more about legal texts. Of The State Decoded’s numerous successful partnerships, the strongest is likely its partnership with the OpenGov Foundation, which has led efforts to implement The State Decoded in Maryland, San Francisco, and Baltimore. Other notable partners of The State Decoded include the ReInvent Law Laboratory at the Michigan State University College of Law, which has committed to implementing the State Decoded in Michigan, and the Free Law Project’s CourtListener, which gathers court decisions online to support The State Decoded for CourtListener’s API. The project has met its key development goals, and as planned from the start, Jaquith has moved on from the project. But the open source community continues to refine and develop The State Decoded’s codebase, and the task of deploying the platform in additional states and municipalities around the country will be left in the hands of motivated and engaged citizens in each community. REACH AND OUTCOMES

By March 2014, The State Decoded had launched in Virginia, Florida, Maryland, Chicago, San Francisco, Baltimore, Washington DC, and Philadelphia. Stakeholders in Delaware, Louisiana, Michigan, and Washington have also expressed interest in implementing it. Accessing copies of laws for other states has proved more difficult than the team anticipated, as most states do not provide bulk downloads of their

laws. Jaquith’s goal was to create a platform that could be applied to all 50 states, spearhead implementation in a select few states, and trust the remainder of the work to volunteer open government groups and citizen activists, with Jaquith serving in a largely advisory capacity. That scenario is exactly what has happened: Jaquith’s team has overseen creation of the Virginia and Florida State Decoded sites, with the OpenGov Foundation and other groups managing sites in other areas. In March 2014, the OpenGov Foundation launched americadecoded.org as a central directory for all Decoded sites in the United States. The team does not track visitors to www.statedecoded.com or any of the individual Decoded sites, as it has prioritized having an API and machine-‐readable data. Nonetheless, there is evidence that the project has had a profound impact on attitudes about open government. Jaquith has become one of the most recognized players in the open government movement, regularly receiving speaking invitations from think tanks and conferences related to freedom of information. He was named a 2011 White House Champion of Change. In 2013, with further support from the Knight Foundation, he spearheaded the foundation of the US Open Data Institute to encourage collaboration among people, organizations, and businesses working on open data. Additionally, in contrast to three years ago, the open data community widely sees state codes as primary datasets that need to be online, and there is widespread interest in making legal codes machine-‐readable, understandable to non-‐attorneys, and automatically integrated with other forms of related legal data. According to members of the open government community, Jaquith and The State Decoded are nearly single-‐handedly responsible for initiating this culture shift.


2010 KNC Winner Profiles


Basetrack 63

City Tracking 66

Front Porch Forum 69

Game-O-Matic 72

LocalWiki 74

Now Spots 77

OpenCourt 79

PRX Story Exchange 82

SeedSpeak 84

SocMap 86

Stroome 88

Tilemill 90

Interim Review: 2010 Knight News Challenge Winners 1

Basetrack (One-Eight)


Basetrack November Eleven An online journal and social media resource center providing continuous coverage of the entire deployment of a US Marine battalion to southern Afghanistan

$202,000

Basetrack created an independent, civilian online journal and social media resource center that provides continuous coverage of the deployment of the First Battalion, Eighth Marines (1/8), to southern Afghanistan. Military media units provide some reporting for the Marines, but their reporting is not independent and the reporters often lack knowledge about the areas in which they are based. This means that the families and friends of military members can only expect passing and superficial reports about those deployed to Afghanistan. Basetrack aimed to provide a platform for reporting to and drawing reporting from the Marines and their families, in order to broaden the perspectives that surround US military operations and to better inform the Marines, their families, and the public. THE INNOVATION

Basetrack’s site combined original reporting from a network of embedded reporters in Afghanistan with aggregated news and analysis about wider regional issues and user-‐generated content (photos, video, and commentary) from the Marines themselves. Its WordPress-‐based platform integrates existing popular social media products (such as Flickr, Vimeo, Twitter, and Facebook) to host and broadcast content created by Marines and by the project’s embedded reporters. Basetrack allowed the troops and their families to be interactive audiences: they steered, challenged and augmented coverage of the 1/8 Marines’ deployment in Afghanistan, and distributed content through their own social media channels.

IMPLEMENTATION

The project aimed to chronicle new uses of social media by the military. Basetrack originally intended to employ existing social media frameworks such as Twitter, Facebook, Vimeo, and Flickr through a relatively simple platform which would require very little software development. Ultimately, however, Basetrack decided to create a more complex platform where content was organized by, and posted in relation to, its location on a web-‐based map. Basetrack struggled to find reliable, effective programmers to develop its WordPress-‐based publishing platform, and the platform ultimately suffered from various technical glitches that made it difficult to load pages, navigate, and view posts. As a result, activity on Basetrack’s primary website subsided and much of the activity took place on its Facebook page. To gain an audience among 1/8 Marines and their family members, the project relied entirely on word-‐of-‐mouth and viral peer-‐to-‐peer marketing through social media channels.

Basetrack’s first team of embedded photo journalists traveled to Afghanistan in September 2010. The project had originally planned to host three to four full-‐time contributors, but because of the danger inherent to being embedded, it instead ended up using one full-‐time staff member and more than a dozen part time-‐time contributors. Basetrack’s embedded photo journalists documented the Marines’ daily operations through essays and photographs taken with the iPhone’s Hipstamatic Application.



0

100

200

300

400

500

Jan-‐11 Mar-‐11 May-‐11 Jul-‐11 Sep-‐11 Nov-‐11 Jan-‐12 Mar-‐12 May-‐12 Jul-‐12

Number of Social Media Mentions

blogsearch facebook twi\er

In addition to the difficulties the project faced in working with outside developers, Basetrack encountered a number of other significant technical and operational challenges. Gaining internet access in remote areas of southern Afghanistan was a challenge and required the team to use expensive and often ineffective satellite data modems. The project also encountered resistance and an unusual level of restrictions placed upon its embedded contributors by the battalion’s commanding officers. In response, Basetrack created redaction software designed to foster radical transparency of the military’s censorship policies. The software allows commanders to black out any text or images, but requires that they indicate that the item was censored, provide an explanation, and assign an officer to be held responsible for the censorship. The software made it easier for the military to

censor Basetrack’s content, but helped the project avoid unnecessary censorship when commanding officers found explaining their redactions too awkward.

Basetrack’s embedded contributors were asked to leave six weeks ahead of schedule on February 5, 2011 amid concerns about the mapping tool’s “perceived Operational Security violations.” Follow-‐up e-‐mails from the military concluded that “media ground rules were not violated” and a definitive explanation for the reasons for terminating Basetrack’s access was never given. The cancellation notice was issued shortly after Basetrack was given an ultimatum to shut down its Facebook page.

Following the ejection from their embed in Afghanistan, Basetrack’s project team went on to present on the project at conferences around the world and to build wireframes for a future

0 5,000

10,000 15,000 20,000 25,000 30,000 35,000 40,000 45,000

Aug-‐11 Sep-‐11 Oct-‐11 Nov-‐11 Dec-‐11 Jan-‐12 Feb-‐12 Mar-‐12 Apr-‐12 May-‐12 Jun-‐12 Jul-‐12

Daily Engaged Users Daily Total Reach

Number of Facebook Users Reached



redesign of its website. Basetrack produced a book of the images and the Julliard School in New York City adapted Basetrack’s photos, videos, essays, Facebook transcripts, and tweets into a multi-‐media live performance. REACH AND OUTCOMES

Though it was asked to leave its post in Afghanistan prematurely, Basetrack was successful in building an audience of the US First Battalion, Eighth Marine Corps members, their families, and other members of the public. Analytics on the project’s Facebook page (which hosted the majority of the project’s activity after the 1/8 Battalion returned home from its deployment in Afghanistan) suggest that its daily total reach averaged nearly 27,000 individuals per month from August 2011 until July 2012. During the time when the 1/8 marines were deployed in Afghanistan, Basetrack’s website was updated with 238 posts, 1460 images, and 100 embedded videos. Seven embedded reporters contributed to the primary content of the site, in addition to the Marine officers who were in charge of redaction.

Basetrack provided an invaluable benefit to military family members, connecting them to information on their loved ones, to other military families, and to background information on Afghanistan. Evidence of Basetrack’s impact can be found in the direct communication from family members of 1/8 Marines via postings to Basetrack’s website and Facebook wall. Although independent-‐minded, vocal, and engaged, the parents and friends of the 1/8 service members were also respectful and civil, expressing very little hostility in their Facebook postings. Users shared articles of interest and generally engaged in intelligent debate.

Essays and photographs from the project have been used in a number of well-‐known outlets including the New York Times, Foreign Policy magazine, Newsweek, BBC/PRI’s The World, Wired, and Gizmodo. Basetrack was the

recipient of a number of awards, received widespread media attention from the traditional press and blogs, and generated significant conversation in online forums and social networks (both positive and negative). Project leader Teru Kuwayama has also been invited to speak to Afghanistan-‐bound US military forces and provide advice on information operations strategy. The project team aimed to use Basetrack as a replicable model that could be imitated in other military units to provide an in-‐depth, wide-‐spectrum view of US military operations. Other military units have expressed interest in potentially hosting a similar project, but it looks unlikely that Basetrack will return to Afghanistan. Its project team is developing and upgrading its WordPress software piece with the intention of making it usable for any kind of blog or media project.



CityTracking


CityTracking Stamen Design, LLC A web service and open source tools to display public data in easy-‐to-‐understand, highly visual ways

$412,000

Federal, state, and municipal governments are publicly releasing previously unavailable datasets. There has not been a corresponding increase, however, in the tools that make that data legible to the public. CityTracking is a project of Stamen Design, a San Francisco-‐based design and technology studio, which aimed to create open-‐source toolsets to present and host urban information in a manner that was easy for technically adept journalists, cities, and the public to use. The project created web-‐based tools which would allow users to create highly polished, easily sharable maps from public data. Ultimately, the project wants data visualization to become part of the core of local information offered by cities, civic groups, and local businesses. THE INNOVATION

CityTracking is a public project comprised of open-‐source toolsets for presenting digital data about cities that journalists and the public can easily access and use to create highly polished maps. Although there are a number of other web-‐based platforms for creating maps (such as Google Maps, OpenLayers, Polymaps, and Tableau), the project’s main innovation was to raise the bar on the visual appeal of easily creatable maps, turning map making into an aesthetic and cultural exercise. Through CityTracking, Stamen designed unique and aesthetically pleasing cartographic styles based on data from Open Street Map. CityTracking’s cartographic styles include:

• Toner: A high contrast, minimalist mapping style which uses only black and white to create a basemap.

• Terrain: A mapping style which includes shaded hills, larger text, and green for park

areas as an open-‐source alternative to the terrain style of Google Maps.

• Watercolor: This style which incorporates colorful textures that appear to be hand-‐painted.

• Burning Map: The burning map style uses fiery animations to represent streets.

• Trees, Cabs, and Crime: A mapping style only available within San Francisco that represents the datasets for tree locations, taxi cab GPS positions, and crime reports in colorful halftones.

IMPLEMENTATION

Stamen Design won Knight News Challenge funding to create tools to help make data visualization part of the core of local information offered by cities, civic groups, and local businesses. The goal was to change the way people view, talk about, and use digital city services. At the outset, CityTracking planned to market its tools to cities, journalists, and the public in San Francisco, Oakland, and Berkeley, and then expand to other cities. As part of the project, CityTracking also planned to host an annual Data Visualization for Cities Conference which would build interest in CityTracking’s tools and explore the current state of the field. The project also planned to regularly add code to CityTracking’s open source code base through GitHub to allow other groups and developers to use its code to build server-‐side data visualization programs.

The project released the following open source toolsets, among others, through City Tracking:

• Dotspotting: Dotspotting is a hosted web service that allows cities and citizens to



upload geographic information, host it, and embed and export it in various ways. The project was made available to the public on December 7, 2010 and formally launched on June 17. Cities including San Francisco, New Haven, and Los Angeles used Dotspotting to map a variety of their urban datasets on crime, vehicle crashes, and the locations of prisons. The tool allows users to place a variety of different types and colors of dots on a map.

• Toner: CityTracking released a second open-‐source tool, Toner, to the public on June 29, 2011. Toner is the underlying code component of the distinctive black-‐and-‐white mapping style used in Dotspotting.

• Flamework: Flamework is an open source framework for web applications that CityTracking created for Dotspotting. Flamework continues to be actively developed as a part of the Dotspotting project, and has been used in a number of Stamen Design’s commercial projects including work for MapQuest.

• Tile Farm: CityTracking spent the majority of its effort on Tile Farm, available at http://maps.stamen.com. Tile Farm is an open source tile generating engine which provides users with a browsable, embeddable, and otherwise immediately usable map of the whole world which can be used in Google Maps, Modest Maps, and other mapping APIs without having to

download OpenStreetMap or needing to work with servers and technical code.

After publically releasing Dotspotting, CityTracking realized that users expressed more interest in the tool’s background maps than in its ability to put dots on a map. CityTracking responded by making the code for its Toner style available for download and by creating Tile Farm, the open-‐source map of the world which can be used to style and download highly polished street-‐level map tiles. CityTracking hosted its inaugural Data Visualization for Cities conference at Stamen’s headquarters in San Francisco's Mission District. The conference gathered over 40 city officials, data visualization experts, and people working in tech. It featured a mix of talks by Stamen personnel demonstrating their work and panel discussions and workshops in which city employees and data visualization practitioners shared their experiences.

Despite the interest the conference generated in CityTracking’s tools, marketing proved to be the largest challenge for the project and CityTracking struggled to find the capacity on its 12-‐person team to provide effective outreach and training to potential users. Without a team member in charge of marketing and publicity, the project heavily relied on industry speaking engagements and Stamen’s blog posts and tweets to build awareness and adoption of its tools. CityTracking also found that while it primarily aimed to reach journalists and the

0

10,000

20,000

30,000

40,000

50,000

60,000

70,000

Jun-‐12 Jul-‐12 Aug-‐12 Sep-‐12 Oct-‐12

Pageviews of Maps.Stamen.com

Pageviews Linear (Pageviews)



public, the majority of its users were the developers and designers working for journalists and cities. REACH AND OUTCOMES

CityTracking experienced less adoption during its grant period than the project team had hoped but has nonetheless made an impact, providing tools to help people gather cities’ data and make that data legible through mapping and visualization. The number of embedded maps created using the project’s first mapping application, Dotspotting, is unavailable, but the average monthly traffic for unique visitors to Dotspotting’s website shows an average of 810 unique visitors a month for 2011, and 749 unique visitors a month for 2012. The use of at maps.stamen.com is significantly higher, with a total of 116,579 embeds between its creation and November 2012.

As noted above, CityTracking found that the majority of its users were developers and designers working for journalists and cities. A number of cities across the United States have used CityTracking, including San Francisco, Oakland, Berkeley, Los Angeles, and New Haven. The state government of California and the national government of Bosnia have used CityTracking as well. The project’s blog provides several examples of how individuals are adapting and using their maps in creative ways: using watercolor maps as art, depicting maps on sneakers, mapping health inspections in Atlanta and wildfires in Colorado. CityTracking also noted the positive, enthusiastic response it

received from cities such as San Francisco and its mayor’s Office of Economic Development.

An important goal of CityTracking was to build an open source code base which would allow other groups and developers to use its code to build server-‐side data visualization programs. From July to September 2011, there were a total of 11 public data repositories and 6 unique contributors to CityTracking’s code on GitHub. From September 2011 to July 2012, these numbers increased only slightly with 15 public data repositories and 10 unique contributors. By August 2012, CityTracking’s code was forked 341 times, 558 people had identified bugs, and 837 people had contributed changes that were incorporated into CityTracking’s code. Both CityTracking and Dotspotting are being closely watched by influential members of the open source community on GitHub.

CityTracking has experienced considerable growth in the two years following its Knight News Challenge grant period, gaining hundreds of thousands of users and averaging over 5,000 embeds per month. The project team attributed this growth to their sustained promotion efforts and their tools’ ease of use.

Embeds from Maps.stamen.com

Cartographic Style Number of Embeds

Watercolor 63,686

Toner 42,620

Toner-‐labels 8,262

Terrain 1,409

Toner-‐lines 373

Toner Lite 229

Total embeds 116,579



Front Porch Forum

PROJECT GRANTEE INNOVATION INVESTMENT

Front Porch Forum

Front Porch Forum, Inc.

A network of online neighborhood forums in Vermont that allow users to read and share posts with their neighbors

$220,000

This project involved the scaling of Front Porch Forum, a mission-‐driven, for-‐profit business that hosts networks of local online forums. Front Porch Forum (FPF) offers an easy-‐to-‐use online platform for communicating with neighbors and keeping up with neighborhood news. The project received convertible debt financing from the Knight News Challenge to: 1) further scale the work of its 25 pilot towns by rebuilding and enhancing Front Porch Forum’s proof-‐of-‐concept software, and 2) expand to cover each of Vermont’s 251 towns.

THE INNOVATION

Front Porch Forum was created to help users meet and get to know their neighbors. By circulating daily neighborhood postings on topics ranging from block parties and lost pets to local politics, Front Porch Forum aims to better inform users about nearby goings-‐on, strengthen a sense of offline community, and spur civic engagement. Front Porch Forum was the first project to enter the online space of “helping neighbors connect,” and since its launch, over 20 groups have started similar projects. Many of these projects, such as NextDoor.com, appear to have been significantly influenced by Front Porch Forum’s code and success.

IMPLEMENTATION

Front Porch Forum’s pilot had been operating for three years, and was already running in 25 northwest Vermont towns, before the Knight News Challenge award. After the award, FPF used an outsourced tech team to rebuild its web application via Ruby on Rails—an open-‐source, agile web application development

framework. It then launched the new web application as the open-‐source OpenPorch on GitHub. In July 2011, FPF also launched a redesigned website. The platform is free of charge to users and allows them to submit postings over email or through FPF’s website. FPF employs online community managers who organize and moderate these postings, stopping negative and recursive threads and ensuring a reasonable balance of content from neighbors/residents and local public officials. To help foster a greater sense of offline community, each posting includes the member’s full name and street name. Registered members receive these postings through daily e-‐newsletters and can access past newsletters through the archives on FPF’s website.

Eager to expand throughout Vermont and beyond, the project developed a marketing plan that project director Michael Wood-‐Lewis described as “complex, authentically local, and relentless.” FPF focused its marketing efforts on partnering with local groups, including municipal governments, nonprofit organizations, chambers of commerce, school districts, and other institutions that would market the project to their employees and constituents in exchange for FPF access and ad space. The project also worked to earn media coverage on its expansion and to place subscriber success stories through local newspapers, radio, TV, websites, and newsletters.

FPF has spread around the periphery of its existing communities largely through word of mouth. To keep up with this growth, its platform is currently undergoing another round of development aimed at building out



components of the software that will further facilitate scaling. A number of towns have approached Front Porch Forum, requesting to launch the platform in their communities. In response, FPF has since changed its business model to require a start-‐up fee for launching into new areas. Communities have paid this start-‐up fee through their chambers of commerce, citizen fundraising, or municipal budgets.

REACH AND OUTCOMES

Since receiving funding through the Knight News Challenge, Front Porch Forum has spread from 25 to 84 towns, including 82 towns in Vermont and one town each in New Hampshire and New York. By reaching 82 towns in Vermont, FPF has achieved about 33 percent of its long-‐term goal for scaling. The project has also spread to the neighboring communities of Stewartstown, New Hampshire and Argyle, New York. One of the key metrics used to measure Front Porch Forum’s adoption is its “take rate”—the percentage of registered users within a given FPF neighborhood. As of October 2012, Front Porch Forum’s take rate was 38 percent, with 43,000 total members out of a coverage area that encompasses 112,000 households. The project’s take rates within individual communities vary between 15 percent and over 90 percent, in communities where multiple registered FPF users exist within the same household.

The project also shows strong signs of user engagement. In communities such as Burlington, more than half of FPF’s users actively post to their neighborhood forum. From July 2011 to July 2012, time spent on Front Porch Forum’s site averaged nearly five minutes (4:50) across 1.5 million page views, with users accessing an average of 5.7 pages of content per visit (even though most users interact with their local FPF via email rather than the website). Most notably, Front Porch Forum’s number of returning visitors over the 12-‐month period was 57.4 percent, which substantially exceeds the industry average. The project’s mentions through Blogger, Facebook,

and Twitter have been growing steadily since November 2011.

Front Porch Forum’s ultimate goals were to help inform users about local news, strengthen a sense of offline community, and spur greater levels of civic engagement. Evidence of these impacts exists on FPF’s blog (which has been maintained for over five years and includes over 1,500 posts), through the thousands of posts made weekly to the project’s forums, and through the outpouring of praise and thanks from users who feel more connected, informed, and involved.

Front Porch Forum has proven to be a powerful tool for community development and building social capital. In the aftermath of Hurricane Irene in late August 2011, FPF was invaluable in broadcasting messages from public officials and in helping devastated Vermont communities coordinate relief efforts. Smaller towns used

Front Porch Forum to post ads seeking emergency housing and volunteers with trucks and chains who were willing to help pull cars from flooded areas. The example of Moretown provides a useful case-‐in-‐point. After the



0

1000

2000

3000

4000

5000

Q1 2010

Q2 2010

Q3 2010

Q4 2010

Q1 2011

Q2 2011

Q3 2011

Q4 2011

Q1 2012

Q2 2012

Q3 2012

Registered New Users

Registered New Users Linear (Registered New Users)

0

5,000

10,000

15,000

20,000

25,000

Returning Visitors New Visitors Linear (Returning Visitors) Linear (New Visitors)

New vs Returning Visitors

hurricane hit, a group of students decided to offer their volunteer services to ravaged communities. The students traveled from town to town, offering their services. Towns that had not been using Front Porch Forum often were unable to put the volunteers to good use. The volunteers would arrive, ready to help, but residents were insufficiently organized to provide them with meaningful work to do. But residents of Moretown— who had been using Front Porch Forum for a year—knew exactly how they could use the volunteer assistance and had the community networks in place to put them to immediate use.



Game-O-Matic (The Cartoonist)


Game-O-Matic Georgia Tech Research Corp.

A free, easy-‐to-‐use tool that allows journalists to build cartoon arcade games based on their news content

$378,000

In partnership with the University of California Santa Cruz, the Georgia Institute of Technology received News Challenge funding to create a free tool that allows anyone to create cartoon-‐like current event games that can be easily integrated into the websites of local newspapers and media outlets. The aim of Game-‐O-‐Matic (formally called The Cartoonist) was to increase the use of news games to convey editorial opinion, helping journalists and editors draw communities to their local newspapers, and further inspire citizens to explore the news. THE INNOVATION

The developers conceived Game-‐O-‐Matic as a free tool for those without a background in game development to use in generating simple, cartoon-‐like current event games that are the equivalent of editorial cartoons. Several other projects exist to develop digital games to build citizen engagement in important issues.1 The expense, time, and expertise required to craft regular video-‐game content, however, has prevented the widespread adoption of news games by traditional media sources. Game-‐O-‐Matic was created to relieve the burden of programming and design while encouraging journalists to think of news events not just as stories, but as systems that can be modeled and explored. The tool serves as an intelligent operating system for creating arcade-‐style games through the process of concept mapping relevant actors and their relationships. By answering a series of questions about the major

1 Other new games-‐focused projects include 2007 Knight News Challenge Winners New York News Games, Oakland Jazz & Blues Clubs Video Game, and Playing the News.

actors in a news event and making value judgments about their actions, Game-‐O-‐Matic automatically proposes game rules and images. Once created, users can publish their games to Game-‐O-‐Matic’s site or to their own website or Facebook profile.

IMPLEMENTATION

Because Game-‐O-‐Matic set out to create a technology from scratch, the project spent its early months conducting research into game design platforms and working to find ways in which the platform could interpret user input to generate games. After a conducting survey of available game design platforms, Game-‐O-‐Matic chose to use PushButton Engine, an open source framework for building Flash games. The project spent a significant amount of time developing a theory of meaning and rhetoric for two-‐dimensional, arcade-‐style games. No one had undertaken this research before, and the development of Game-‐O-‐Matic took longer than the project team anticipated since they had to conceptualize the types of stories the tool could process, the basic structures of the news, how the tool could combine video-‐game elements to create meaning, and how to make the software usable.

Game-‐O-‐Matic’s project team programmed an early version of the tool after they determined the video-‐game elements that one would need to portray events, coded ways these pieces could be combined, selected a method by which events are broken down into actors and relationships, created interpretations of relationships that can suggest nuance in a story, and coded an interface that allows journalists to input stories. As the project rolled out an early



version of Game-‐O-‐Matic, called the The Cartoonist, it received a number of angry messages and comments from editorial cartoonists who viewed the tool as way for newspapers to replace them. As a result, the project was forced to reconsider its branding and to spend time explaining that it intended the tool to reference editorial cartoons’ ability to convey bite-‐sized commentary on current events rather than replace cartoons entirely. The project ultimately chose the name “Game-‐O-‐Matic” to convey its automated process for creating games and the complex, generative nature of the tool.

The project intended to market Game-‐O-‐Matic to the editors, reporters, and designers of local newspapers and media groups. However, later on it expanded its target audience to include the general public in order to create greater adoption and awareness. Game-‐O-‐Matic reached out to local news organizations in Atlanta, GA and Santa Cruz, CA to test the tool and inform its development. It also approached individuals from other local news organizations and presented on the project at various game developer conferences and to the World Newspaper Congress and World Editors Forum in Vienna, Austria. Many of these target groups were initially receptive, but Game-‐O-‐Matic found it difficult to promote a novelty product, and the project struggled to find highly-‐visible partners who would create a game to serve as an example for other potential users. The project developed a beta site, http://game-‐o-‐matic.com/, but is waiting to publicly launch until the tool is more polished and easier to use. REACH AND OUTCOMES

Game-‐O-‐Matic has yet to launch publicly and is still in the very early stages of its lifecycle. By the end of the grant period, adoption of the tool had been low: by August 2012, 450 games had been produced and news media groups were experimenting with the tool. Anecdotally, user reception of the tool was largely positive, but Game-‐O-‐Matic found it difficult to motivate these partners to actually use the tool to create games that they posted on their websites. As of

mid-‐October 2012, Game-‐O-‐Matic was still struggling to find a user-‐created game to serve as an example of the tool’s use to promote to local newsgroups. Overall awareness of the tool is also low. Social media monitoring captured about 50 mentions of “Game-‐O-‐Matic” between March 2012 and November 2012, with mentions spiking around times when the project gave presentations during conferences and events, including the Game Developer’s Conference, Games for Change, and the workshop on Procedural Content Generation in Games.

The project’s ultimate aim was to create a tool which would increase the use of news games, resulting in readers’ increased interest in the news. To date, there is no evidence to suggest that Game-‐O-‐Matic has significantly helped to increase the use of news games in local news or raised reader interest in the news. Perhaps the project’s greatest achievement, however, was the successful completion of a beta version of its platform. Game-‐O-‐Matic proved itself to be a versatile way of thinking about meaning and games and even produced a white paper on the theory of meaning and rhetoric for two-‐dimensional, arcade-‐style games.2 Its model established a flexible framework which can be updated with new templates, variables, assets, and logics, and could inspire the design and creation of other news game authoring systems in the future.

Game-‐O-‐Matic’s future plans include pursuing bridge funding of $50,000 for the next year in order to revise the tool according to initial users’ feedback, to eventually launch the platform to the public, and to continue pursuing partners who could market games made with the tool to other potential users.

2 Game-‐O-‐Matic’s white paper is available here: http://mtreanor.com/research/micro-‐rhetorics.pdf



LocalWiki


LocalWiki WikiSpot An easy-‐to-‐use, open-‐source “wiki” platform tailored to the needs of local communities

$360,500

Wikis are websites developed collaboratively by a community of users, allowing any user to add and edit content. Based on the success of its first locally based wiki in Davis, California, LocalWiki received Knight News Challenge funding to create specialized, open-‐source wiki software and to help other geographic communities develop, launch, and sustain wiki projects using this software. THE INNOVATION

LocalWiki pioneered locally focused wiki software with the goal of making it easy for people to share knowledge of their own communities. The platform enables users to create pages and articles, upload photos and files, and edit others’ pages. Community wiki pages can feature articles on anything from the news and history of the area to updates on local public transportation, nearby attractions, lost pets, or local social services. Specific innovations within LocalWiki’s software include its accessibility and ease of use, tagging features for pages, and mapping capabilities that allow users to link pages to points and areas on a map. Users can easily view and reverse edits to pages and maps and track their wiki’s number of pages and contributors through a dashboard feature. IMPLEMENTATION

LocalWiki set out to develop an enhanced, open-‐source wiki platform with documentation designed to meet the needs of local communities. Because the project’s Knight News Challenge funding focused more heavily on software development, LocalWiki was limited to launching and testing its software in

just one pilot community. LocalWiki turned to the crowd-‐funding platform Kickstarter to help fund outreach and education efforts and eventually raised over $26,000 with the help of 427 individual donors. Delays in the initial disbursement of Knight grant funds pushed back the start of the project by about three months, but progress picked up when the project received its funding. LocalWiki publicly released its first version of the software on November 30, 2011. The platform was designed to be easier for non-‐developers to install and to create a more accessible editing process. LocalWiki marketed this software to the open-‐source community through the project’s own development mailing list, through interaction with other projects on GitHub, and through a series of hackathons on LocalWiki’s code.

LocalWiki selected Denton, Texas to serve as its first pilot community because Denton’s project leads seemed the most able and motivated to work under limited initial guidance. By piloting DentonWiki, LocalWiki aimed to gain essential feedback on the new software as it was being developed, and to gain experience helping other communities create and launch local wiki projects. Prior to the launch of DentonWiki, LocalWiki worked with the team in Denton to



help build a core, dedicated group of editors who would drive the project. Denton’s wiki featured over 800 pages when it launched, on the same day as the public release of LocalWiki’s new software.

After supporting the first pilot community in Denton, Texas, LocalWiki selected Santa Cruz, California and Raleigh-‐Durham, North Carolina as additional communities in which to test the platform. LocalWiki left the local outreach work up to these focus communities but offered them coaching along the way through visits, phone calls, and emails. By providing direct support to a few successful communities and then building a network of communities implementing the software, LocalWiki hoped to create a model in which communities would reach out to each other for best practices and further spread the platform.

REACH AND OUTCOMES

LocalWiki is well positioned to achieve its ultimate aim of demonstrating that local wikis are a viable way for communities to manage and share information. LocalWiki’s open source code has been adopted and used around the world. Between late August 2011 and the software’s first major release in late November 2011, LocalWiki’s platform was installed nearly 700 times. LocalWiki is now the second-‐most installed of the Knight News Challenge’s server-‐focused projects.1 By late August 2012, there were at least 58 independent LocalWiki projects (of which 37 were considered “major”), spread across nine countries and published in seven languages. LocalWiki only hosts a minority of its community projects, but it received about 26.7 million page views for these projects over the

1 LocalWiki’s downloads are second only to the Ushahidi platform, a server-‐focused project which provides an open source software for collecting, visualizing, and mapping citizen reports from large news events.

0 100 200 300 400 500 600 700 800 900 1000

Independent Installs

0 5,000,000 10,000,000 15,000,000 20,000,000 25,000,000 30,000,000 35,000,000 40,000,000 45,000,000

Content Growth (in pages)



course of 2011.

By late August 2012, LocalWiki’s code had been downloaded over 690 times and forked 40 times. It had also attracted 164 individuals who signed up to be notified of changes to the code. Additionally, 73 people have contributed by identifying bugs in LocalWiki’s code, and LocalWiki has incorporated others’ changes in about 80 cases. Users have downloaded the code to develop mobile web applications, a LocalWiki-‐like project for Burning Man, and new functionalities such as editable comments. Two users in Portugal used the code to internationalize LocalWiki, so that the platform’s user interface could be translated into any language. To date, LocalWiki is available in seven languages. Although the platform has exerperienced impressive adoption, most of LocalWiki’s community wikis are still working to reach a critical mass of users and become entirely community maintained. LocalWiki has already reached this tipping point with its longstanding DavisWiki project. Nearly one in seven Davis residents contributes to the wiki at some point within the course of a year, and about half of Davis’s residents use DavisWiki each month. LocalWiki’s other communties have yet to achieve this same level of engagement but have demonstrated a number of initial successes. Santa Cruz, California is LocalWiki’s largest focus community, with over 5,400 pages, 4,300 photos, and 2,700 maps. TriangleWiki in Raleigh-‐Durham, North Carolina has seen widespread adoption and use from Raleigh’s city council, which has used the wiki to post city government content. Two locally based mobile projects have also used TriangleWiki, incorporating the wiki’s content streams in their applications.



NowSpots


NowSpots Windy Citizen Open-‐source software allowing “real-‐time advertising” that can be updated at any time by local businesses using social media

$257,500

NowSpots was created to provide local publishers with “real-‐time advertising” through widgets that show the latest updates from an advertiser’s social media accounts. The project aimed to turn sponsored social media streams into a viable source of income for news sites. THE INNOVATION

Traditional web advertising relies heavily on banner ads—static or animated images that display an advertiser’s messaging each time the page is refreshed. If done well, banner ads can build brand awareness and help drive traffic to the advertiser’s site. For smaller organizations without a strong web presence, however, these ads often fail to build a relationship between viewers and the organization being advertised. NowSpots created open-‐source software allowing “real-‐time advertising” through a widget that shows the latest message or post from the organization’s social media accounts including blogs, Facebook, Twitter, and YouTube. The platform enables users (local businesses and publishers of all sizes) to update their advertisements automatically with real-‐time information, allowing them to connect more directly with potential customers. It also provides them with back-‐end analytics on how many times potential customers view the ads, click on them, or repost content from them. IMPLEMENTATION

The Knight News Challenge awarded funding to NowSpots to release and promote open-‐source code that lets local publishers sell, manage, and serve “real-‐time advertising” on their own sites. The original concept for the project was to

market the tool to local news outlets that could use the tool to add a viable source of income by selling these advertising spots to businesses. To help streamline newspapers’ sales process, NowSpots also developed a search tool that scours local businesses, identifies which are the most active in social media, and rates them on how likely they are to be interested in purchasing “real-‐time advertising.” The project worked to build awareness through an internal sales force that targeted about 400 local newspapers around the country, relying on conversations and word of mouth.

NowSpots partnered with its first client, the Chicago Tribune to test the tool, working with the Tribune’s sales team to market the real-‐time advertising spots to outside businesses. Despite some initial successes with the Chicago Tribune, NowSpots’ momentum eventually waivered and the paper’s sales team stopped selling the advertising spots. NowSpots estimated that the Tribune’s real-‐time ads were performing well, with click-‐through rates nearly 300 percent higher than traditional ads, but it proved to be difficult for the Tribune’s sales team to sell to outside businesses. NowSpots also partnered with a number of other newspapers and media groups and encountered similar challenges in educating the papers’ sales staff about the product and how to effectively sell the advertisements. NowSpots found it difficult to motivate news organizations’ sales teams, as these teams typically had 30–50 other products to sell.

As a result of these challenges, NowSpots pivoted from targeting news organizations to selling the tool to small businesses and start-‐



ups. In early October 2012, NowSpots founder launched Perfect Audience, a Facebook and web retargeting platform that companies can use to target Facebook ad campaigns to people who visit the company’s website, with the aim of helping them reach their ideal customer at scale. REACH AND OUTCOMES

By November 2012, the information needed to determine the extent of NowSpots’s adoption and ultimate impact was unavailable. Measuring adoption of NowSpots ads was difficult, since the ads ran on publishers’ websites. However, by fall 2011, NowSpots had tested its system in 447 news media outlets, including the Chicago Tribune, Hearst Digital media, Gatehouse Media, and Digital First Media, and in a number of ad agencies and small businesses.

Despite the challenges it encountered in working to train news organizations’ sales team to sell NowSpots ads, early users such as the Chicago Tribune believed the project met a

need and customers such as Mastercard, Northwestern University, Hard Rock Café, and the Art Institute of Chicago each bought advertising spots from the project. Some of NowSpots most successful advertising campaigns with these customers have lasted longer than eight months. NowSpots’ six-‐month relationship with the Chicago Tribune alone produced 25 advertising campaigns with over seven million ad impressions. Results from the project’s test markets indicate that NowSpots ads’ click-‐through rates were at .361 percent—about 300 percent higher than the average click-‐through rate of traditional online banner ads.

Perfect Audience—the Facebook and web-‐retargeting platform launched by NowSpots founder Brad Flora—has achieved considerable success since its founding in October 2012. In June 2014, Perfect Audience was bought for $25.5 million by Marin Software.



OpenCourt (Order in the Court 2.0)


OpenCourt Trustees of Boston University

A pilot project to demonstrate how digital technology can increase public access to the courts

$250,000

Mainstream news outlets often lack the resources to adequately cover local courts. At the same time, court policies regarding digital journalism have not changed since standards for video and audio recording were established in the 1970s, and the judicial branch has failed to adapt to new technologies such as web streaming and social media tools like Twitter and Facebook. OpenCourt is a pilot project designed to test new media initiatives within a working court system and to establish best practices that can be replicated and adopted throughout the nation’s court system. THE INNOVATION

OpenCourt grants citizens and professional journalists digital, web-‐based access to Quincy District Court in the greater Boston area. The project provides the court with a WiFi network for use by journalists and bloggers and features a live-‐streaming video of the proceedings from the court’s First Session courtroom and daily archives of courtroom footage. OpenCourt aims to serve as a model for similar efforts to integrate new technology into the courts as a means of improving the public’s access to and understanding of the judicial system. IMPLEMENTATION

OpenCourt evolved from work done by the Media Committee of the Massachusetts Supreme Judicial Court, and it began with a series of meetings in December 2010 to introduce the project and invite feedback from court staff, attorneys, local journalists, and advocacy groups.

Even before its launch, OpenCourt anticipated a number of challenges related to the fear that cameras could influence courtroom behavior and the outcomes of cases. The project faced resistance from a number of stakeholders, including defense attorneys, the District Attorney’s office, and advocates against domestic violence. Concerns included the difficulty of ensuring that the project would strike the balance between public access and an individual’s right to a fair trial and due process, and issues of safety in protecting the privacy of domestic abuse victims. Despite these groups’ opposition to the project, each participated in its development, serving on the working group that met regularly to establish guidelines and policies for OpenCourt’s live-‐streaming courtroom coverage.

OpenCourt also dealt with a number of unexpected technical challenges and trademark issues prior to its site launch. Gaining sufficient bandwidth into Quincy District Court proved to be difficult and delayed the project’s original timeline. After rebranding from its original project name “Order in the Court”, OpenCourt also learned that TruTV (formerly CourtTV) owned this “mark” and a similar domain name.

The project’s website, http://opencourt.us, went live in May 2011. The site allows users to view live-‐streaming court cases, daily archives of cases, and electronic versions of daily court schedules. On its first day of operation, the local District Attorney’s office filed a motion to close access to OpenCourt’s online archives. The motion was denied, but by July 2011, the District Attorney’s office had filed a pair of motions aimed at shutting off the site’s archives



0

50

100

150

200

250


Number of Social Media Men=ons

Other facebook twi`er

and live-‐streaming footage. The Massachusetts Supreme Judicial Court ruled in favor of OpenCourt in March 2012 and established that the state cannot suppress publication or redact footage of public proceedings. Not long after this decision, the District Attorney’s office again sued OpenCourt to stop its plans to begin streaming jury trials. The court ruled in OpenCourt’s favor again in mid-‐August 2012 and allowed OpenCourt to move forward with live-‐streaming jury trials under the preliminary guidelines set forth by its decision in the earlier case. On September 11, 2012, OpenCourt live streamed a jury trial from Quincy District Courthouse for the first time. REACH AND OUTCOMES

OpenCourt has received significant attention in the legal field and in the media. Trend data on visits to OpenCourt’s website are unavailable, but Google Analytics data shows that the project received a total of 122,038 visitors including 70,788 returning visitors between October 2011 and late November 2012. Spikes in the number of OpenCourt’s social media mentions are due in part to the increased media attention during its legal challenges. The project received positive coverage in local new outlets such as the Boston Globe, the Boston Herald, Quincy Patriot Ledger, WBUR, and Massachusetts Lawyer’s Weekly. Links from Boston.com, the Associated Press, WBUR, and the Quincy Patriot Ledger websites each helped to drive traffic to OpenCourt’s website.

OpenCourt also received attention in national sources such as the Legal Law Network and Current, a periodical that covers issues in public media.

The ultimate goal of OpenCourt was to create a set of benchmarks and best practices for digital communications that could be published and shared widely among industry groups, even beyond judicial systems. This broader vision has gradually come to fruition. In addition to the project’s web traffic and media attention, OpenCourt’s team was asked to present at a number of speaking engagements nationally and received numerous requests from legal journals that are interested in publishing the project’s findings. OpenCourt is in the process of authoring these articles.

Perhaps OpenCourt’s greatest achievement is the precedent that it has set on the issue of citizen access to courtrooms—a precedent that could encourage and strengthen similar efforts in Massachusetts and across the United States. OpenCourt’s legal victories have also had a significant impact on Massachusetts’s revised guidelines for cameras in courtrooms. At the time of this report, OpenCourt had just launched its coverage from a second courtroom, using a multi-‐camera setup to document new types of hearings, such as jury trials. OpenCourt’s broadcast on September 11, 2012 marked the first criminal trial ever to be live streamed in Massachusetts. According to Google Analytics, the trial’s live stream received



a total of 481 views.

OpenCourt’s legal battles, however, did slow its expansion and limited the amount of time the team could devote to developing contextual reporting, training citizen journalists, and developing technical solutions to handle the large amount of video data captured in court. In spite of this, the court cases helped to generate more awareness and interest in the project.

Despite the project’s momentum, OpenCourt is struggling to find a funding model to sustain itself. OpenCourt’s team has funding and institutional support from WBUR and Boston University to keep the project running through December 2012 and is currently exploring a number of potential options for sustaining the project.



PRX Story Exchange (StoryMarket)


PRX Story Exchange

PRX, Inc. A crowd-‐funding platform that allows local public radio stations, producers, and listeners to find and help fund stories

$75,000

Faced with capacity constraints and the high cost of producing local stories, many of today’s public radio stations fail to cover local news. Story Exchange, run by PRX, a nonprofit public media company in Cambridge, Massachusetts, was created to offer an open-‐source crowd-‐funding platform for finding, supporting, and distributing local public radio stories. PRX partnered with an existing crowd-‐funding project, 2008 Knight News Challenge Winner Spot.us, to serve as Story Exchange’s technical platform. The project was funded through the Knight News Challenge to develop a pilot with Louisville Public Media (LPM) in Louisville, KY and to eventually expand to at least three other test markets. Ultimately, Story Exchange failed to gain widespread adoption. THE INNOVATION

Story Exchange aimed to help crowd funding become an effective revenue option for professional journalism. Using Story Exchange, public radio stations and independent producers could issue requests for story ideas along with their respective fundraising goals, hosted on the station’s own website, PRX’s website, and Spot.us’s website. The platform offered listeners the ability to vote on story ideas, to pledge financial support to the stories and pitches of their choosing, and to post suggestions of other stories they wanted to hear in the future. Several other web-‐based crowd-‐funding efforts existed, such as Kickstarter and Spot.us, but Story Exchange was the first to focus on funding public radio stories and the first to promote these stories using on-‐air pitches.

IMPLEMENTATION

PRX set out to test Story Exchange’s model with LPM in Louisville, KY and to eventually scale to other communities already connected through PRX.org. Rather than building its own technical platform from scratch, Story Exchange partnered with 2008 Knight News Challenge Winner Spot.us. Story Exchange originally planned to use Spot.us’s code in order to develop a separate project. After lengthy consideration, PRX chose instead to integrate Story Exchange into Spot.us. Coordinating with Spot.us on the scope, specifications, and interface of the platform took Story Exchange longer than expected and delayed the project’s development.

Story Exchange’s launch met with resistance from public radio journalists. The day Story Exchange began on-‐air announcements for the project’s pitches in Louisville, a number of public radio producers, editors, and consultants posted accusations about the potential for conflict of interest and concerns about the project’s lack of editorial control to the Facebook page of the American Public Media Group (APM). PRX had anticipated this reaction at the outset of the project and had put rules in place that capped the amount an individual listener could contribute to a story, created a policy for returning the donations of listeners too close to an issue, and granted partner stations decision-‐making authority over which stories to air and the content of these pieces. Story Exchange confronted stations’ concerns and misperceptions head-‐on in in-‐person meetings and in written explanations, but the



negative attention ultimately hampered the project’s ability to attract partner stations.

While it lasted, Story Exchange’s partnership with Spot.us benefited both groups and contributed to improvements within each project. However, the dynamic changed significantly in fall 2011, when APM acquired Spot.us. PRX continued partnering with APM on Story Exchange, but the members of the Spot.us team who PRX had worked with were placed on other projects or left the company. As a result, Story Exchange lost its developer team and its platform lacked sufficient technical support and attention. Similarly, Story Exchange’s relationship with LPM was largely positive until the project’s key point of contact left the station. While LPM’s staff was enthusiastic and engaged with the project, the loss of its champion within the station left Story Exchange without a key leader who had helped ensure the continuity and direction of the work. REACH AND OUTCOMES

PRX Story Exchange struggled to gain adoption among public radio stations. It even had trouble finding three other public radio stations to serve as the project’s test markets. PRX made pitches to a number of potential partners, but each was unwilling to adopt the platform. St. Louis Public Radio worried the platform might dilute and distract from the radio station’s general fundraising appeals. North Country Public Radio in the Adirondack area of New York State didn’t have the technical capacity to integrate Story Exchange into its website quickly enough. Other stations, such as KALW in

San Francisco, had begun using Kickstarter as a means for funding news pieces.

Despite this, Story Exchange’s partnership with Louisville Public Media accomplished the goal of helping fund original local news pieces. Four of the five story ideas posted to Story Exchange were fully funded by listeners: 92 individuals contributed a total of just over $900 to these stories, and the four stories that received funding were aired on eight different public radio stations. “Is it Time to Get Serious about Coal Ash?” was Story Exchange’s most successful story. The piece aired over each of Kentucky’s seven public radio stations, inspired a local producer to create a 12-‐minute documentary, and won the award for Best News Series from the Indiana Associated Press Broadcasters. One user of Story Exchange felt that the value it added was less about filling the financial needs of a specific story (often, the station may have done the piece anyway), but providing supplemental funding and motivation to follow through, as the station was then beholden to the story’s donors.

Story Exchange may have failed to gain adoption among public radio stations, but PRX learned a number of valuable lessons about the techniques, guidelines, and knowledge needed to ensure a successful crowd-‐funding campaign. PRX chose to discontinue Story Exchange beyond the two years of its News Challenge grant, but it began using Kickstarter in its crowd-‐funding efforts for independent public media.



SeedSpeak (CitySeed)


SeedSpeak Arizona State University

An application with mobile, web, and widget components that provides citizens in local communities an easy way to suggest community improvements to local leaders, volunteer groups, and each other

$93,600

SeedSpeak, formerly CitySeed, is a project led by Retha Hill of the Walter Cronkite School of Journalism and Mass Communication at Arizona State University (ASU). SeedSpeak is an application with mobile, web, and widget components. It allows users to plant suggestions (seeds) in local communities in order to empower other community members, leaders, and volunteer groups to discover their ideas, add to them, and help bring them to fruition. The project was created to give citizens a simple way to suggest community improvements and to increase the number of people who are informed about and engaged in their communities. THE INNOVATION

A number of similar applications—such as EveryBlock, City Works, City Ideas, and SeeClickFix— exist for suggesting community improvements and increasing community engagement.1,2 SeedSpeak is the first of these to offer a low-‐cost or free open source solution specially targeting smaller cities. The project allows citizens to suggest community improvements at the exact location where they see an unmet need or have an idea for a project. SeedSpeak includes web and mobile applications as well as a white-‐label widget to be incorporated into local news sites or

1 City Works and City Ideas are map-‐based applications created as a part of SocMap (formerly known as GoMap Riga), a fellow 2010 winner of the Knight News Challenge. 2 EveryBlock is a winner of the 2007 Knight News Challenge. EveryBlock.com was acquired by MSNBC in 2009 and now operates in 19 US cities.

municipality’s websites can embed so community members can collaborate to improve their community or to report problems. Problems and suggestions can focus on any number of community-‐related issues, such as traffic, recreation, or the use of public space. In a typical case, a user might come across a potential location for a community garden. The person can use SeedSpeak’s mobile application to geo-‐tag his or her idea, linking it to the exact location of the potential garden. Other users view this and other place-‐based suggestions, debate, and take action on their favorite ideas. IMPLEMENTATION

SeedSpeak experienced several challenges which pushed its timeline back by over a year. It hired a local, Phoenix-‐based interactive agency to design, develop, and build its applications. The project team focused its initial energy on user-‐centered design, researching the desires of the Phoenix, AZ community for features in a mobile, idea-‐sharing social network. SeedSpeak conducted interviews with avid social networkers, mobile experts, city officials, leaders of community organizations, and news gatherers, in an effort to understand the needs and goals of potential users and other relevant stakeholders. The research helped inform SeedSpeak's feature set, layout, and design and allowed the outside developers to hammer out a prototype website design. Early user feedback also helped SeedSpeak revise its design plans, de-‐emphasizing the gardening metaphor of planting and growing ideas after testers



confused the goals of the project with actual gardening.

The third-‐party developers produced an alpha version of SeedSpeak’s website in March 2011, but the site required more funding and months of additional work to address design and functionality flaws. Delays in funding due to ASU’s reporting requirements also stalled the project, causing the developers to suspend their work for nearly two months. After changes in similar applications such as EveryBlock and SeeClickFix, SeedSpeak spent more time than anticipated distinguishing itself from other applications. SeedSpeak eventually created a beta version of its site, http://beta.seedspeak.com/, by fall 2011 and the iTunes App Store accepted SeedSpeak’s iPhone application in early August 2012. SeedSpeak contracted with a new developer to work on the Android application the next month.

When its project team initially conceived SeedSpeak, it planned to host the widget version of its application on local newspapers and news websites to allow for more interaction between users and local reporters. SeedSpeak shifted this focus shortly after receiving Knight funding when it found that media organizations were too distracted by budget constraints and other competing projects. The project pivoted instead to city and municipal government sites, promoting itself through direct outreach to local governments, local political leaders, and community groups. SeedSpeak also planned to market the tool through coverage in local news outlets and in the marketing literature, websites, and outreach collateral of supporting foundations, civic groups, public officials, ASU, and local Chambers of Commerce. REACH AND OUTCOMES

As a result of major delays in developing the beta version of the website, applications, and widget, SeedSpeak only began promoting its tool in fall 2012. By late October 2012,

SeedSpeak was working to identify the first community that would test the platform by reaching out to numerous local governments, political leaders, and community groups. It remains relatively early in the project’s lifecycle to assess the platform’s ability to reach some of its more ambitious goals to increase the number of people informed about and engaged in their communities. However, several communities have expressed an initial interest in using the tool. Among these areas are Chandler, AZ; Jerome, AZ; Yavapai County, AZ; and Benton Harbor, MI.

To sustain the project into the future, SeedSpeak has applied for bridge funding through a partnership between the Knight Foundation and the Association for Education in Journalism and Mass Communication, and through ASU. Despite the challenges and delays faced throughout its development, the project leader is optimistic about SeedSpeak’s ability to land contracts with smaller cities. The project continues to evolve, exploring other possible uses, geographies, and partners. SeedSpeak is also exploring the idea of testing its widget in countries with markets that are less crowded with community engagement applications, such as Mexico.



SocMap (GoMap Riga)


SocMap Society Technologies Foundation

A map-‐based social network where users can browse news and engage in civic action through an online local community map

$265,000

The Society Technologies Foundation received funding through the Knight News Challenge to create SocMap, a social platform where users can browse news and engage in civic action through a live, interactive online map of their community. SocMap (originally named GoMap Riga) is a portal application that connects users for interaction, self organization, and the spread of local news. SocMap was tested in Riga, Latvia with hopes of eventually expanding to other cities. THE INNOVATION

The original innovation behind SocMap was a social network where news would be presented in relation to its location on a live, interactive, web-‐based map. SocMap would pull and aggregate community news found on the web, place these stories on its map, and allow users to browse local news and to search and post local events, pictures, and videos. By being integrated with the major existing social networks, users could interact with local news stories, and have their tweets placed on the map. The application also aimed to provide a platform for civic engagement where users could post initiatives to the map (such as suggestions for a community mural, or complaints of a broken fountain), gather signatures from fellow citizens, and bring the initiative to the attention of the local municipality, media, or police. IMPLEMENTATION

The project did not unfold as planned. Although the project team built SocMap and experimented for a full year with various ways

of attracting users and motivating them to post content to the site, SocMap struggled to reach 1,000 users, the platform ultimately stagnated, and the project team decided to adapt. Using the advanced mapping API that it had created, the project team began creating smaller, more targeted applications that let users interact with municipalities and city governments rather than with other members. SocMap had hired a team of creative developers under the original concept for the platform, but the shift in focus required a new team, one that had business expertise. The core members of the project team, who had been engaged on multiple projects unrelated to SocMap, decided to dedicate themselves full time to the project, and SocMap hired individual directors in business, marketing, products, networking, sales, and technology. SocMap marketed the applications that came out of its original platform to municipalities through the website and Facebook page of the new brand “Stakeholde.rs.” As of November 2012, Stakeholde.rs offered four, white-‐label map-‐based applications:

• City Works allows cities to post completed municipal maintenance projects and enables users to suggest improvements.

• City Ideas allows cities to post questions and ideas on a map and solicit citizens’ input. Users can vote on the most popular ideas or make their own suggestions.

• City Growth presents completed and planned city development projects to citizens and investors. Residents can view these projects, comment on them, and share them over existing social networks.



• Map Survey can be used by journalists and municipal governments to create surveys that visualize users’ responses on a map-‐based infographic. Map Survey was used to create HotBills, a citizen-‐populated data map that visualized the differing costs of heating per square meter across Latvia, and SchoolMoney, another citizen-‐populated map that presented the costs associated with Latvian public education.

The project team found that creating smaller, more targeted applications was cheaper and easier to experiment with than creating and managing a large website. Stakeholde.rs is now translating Map Survey into English and designing other map-‐based applications to market to municipalities across Latvia and the European Union. The project’s partnerships with external groups were crucial in helping promote these applications. SocMap partnered with the Municipality Union of Latvia, an association of local and regional Latvian governments, to help attract the municipalities to use Stakeholde.rs’ tools. SocMap also partnered with the Baltic Center for Investigative Journalism to attract publicity and content for the map-‐based infographics created with Map Survey. REACH AND OUTCOMES

As noted above, SocMap’s original platform failed to gain significant levels of adoption and engagement, struggling to reach even 1,000 registered users and to expand beyond users the project team had attracted through their own personal networks.

Although its first platform failed, the project used SocMap’s mapping API to create four more-‐successful applications marketed through Stakeholde.rs. By October 2012, one-‐sixth of Latvia’s municipalities (about 20 out of 120 cities) are using Stakeholde.rs’ applications to engage their citizens. Stakeholde.rs will begin to turn a profit once approximately 40 more cities purchase its tools. In early October 2012, 13 other European cities signed up for free trials of various Stakeholde.rs tools. Among these cities

were Amsterdam and The Hague, Netherlands; Munich, Germany; Warsaw, Poland; Terrassa, Spain; and Dresden, Germany.

Map Survey has emerged as Stakeholde.rs’s most successful application to date. HotBills took only a few weeks to develop, but within a month of its launch in February 2012, the application was used by two percent of Latvia's population (40,000 people) and over 2,500 users had scanned and submitted their heating bills. Between January 1, 2011 and July 31, 2012, HotBills had 262,593 unique visitors and over 1.1 million page views (90 percent of which came from Latvia), making it the largest crowdsourced journalism project in Latvian history. The tool aggregated this data into a visual map which revealed that the cost of heating per square meter differs by up to several times, that neighboring houses could have vastly different costs, and that Latvians do not know how their bills are calculated. HotBills ultimately helped to provide users with an incentive to talk to their landlords about heating prices, to ask for explanations, and to demand adequate answers.

Stakeholde.rs is primarily focused on marketing MapSurvey to neighboring countries in the Baltic region, but is also in the process of translating the application into English, as a number of local governments in other countries have expressed interest in the tool. The BBC, the Guardian, and various Baltic media organizations have signed up for Map Survey’s free trial. By May 2013, Stakeholde.rs apps reached 30 subscription contracts in Latvia and two in Estonia. The company was acquired in May 2013 by investors with experience in SAAS for government institutions.



Stroome


Stroome Stroome An online video editing community which allows users to upload content and collaboratively edit

$230,000

Eyewitness video captured by mobile phones or webcams has rapidly become a key component of news coverage. Stroome’s platform was designed to allow multiple journalists or aspiring journalists to cover the same story and stream their footage to the web, replacing satellite truck technology with an inexpensive online solution. Stroome aimed to provide users with a robust online editing community where they can collaborate, share ideas and tips, and publish accurate, contextual news in real time. Despite challenges in working with developers and in finding supplemental funding, the platform earned praise and has seen growing adoption among users. THE INNOVATION

Stroome is the first online video editing platform to allow multiple users to collaborate on a project. Other online editing sites exist, but each requires users to exchange and transfer video files through other means (such as email) in order to work together to edit a piece. Stroome aimed to transform collaborative video editing into a more efficient process by allowing users to edit content together within their web browsers, add and view uploads with real-‐time streaming, exchange comments through remix or text, and publish finished pieces through their own blogs, websites, and social media channels. IMPLEMENTATION

Prototyped at USC Annenberg’s Program on Online Communities in the fall of 2008, Stroome started small and aimed to iterate, scale up, and eventually roll out to a larger audience of journalists and journalism students. The project

was on its second iteration when it received its Knight News Challenge award in 2010. It used this funding to launch a third iteration of its platform, which would remedy the broken flow and functionality of its previous versions. Stroome hired a top Los Angeles design and user experience firm to partner with the project, held focus groups to gain user feedback for future iterations, decided on a new logo, and worked to find ways to make the platform’s functions for sharing and collaborating on projects more intuitive.

One of Stroome’s largest challenges was finding the appropriate third-‐party developer for the project. The project team identified only one company that had the technical skills and technology to create the necessary video elements for the project. Stroome ultimately created a partnership with this company, but experienced a number of challenges in the process of rebuilding its older Drupal platform from scratch using Ruby on Rails, which ultimately delayed its release. Stroome eventually launched its third iteration at TEDxUSC, in April 2011.

With this launch, Stroome aimed to become an essential tool in the classroom and for journalists in the field, and focused its initial marketing strategy on targeting students who are enrolled at journalism schools. The project used a strong social media campaign over Facebook and Twitter to build anticipation for its launch and to create an initial user base to help populate the site with content. Stroome integrated the tool in the journalism program in the USC Annenberg School for Communication and Journalism and marketed the tool to a number of other journalism and digital media



programs nationally through email, RSS feeds, word of mouth, and presentations at relevant conferences and trade shows.

REACH AND OUTCOMES

Early signs of Stroome’s impact are evident in the number of users and number of videos edited within the site. By late fall 2012, Stroome had been used by nearly 1,300 users in 146 counties, who had created around 105,000 pieces of content. Stroome also garnered attention from a number of notable sources early on in the project, including blogger Keith Shaw, writer of the influential “Cool Tools Happy Blog,” BBC news web reviewer Kate Russell, and the Guardian, which endorsed it as one of the “top five social networks worth a browse.” To date, the project and its team have been featured in over 150 media outlets and recognized by major media organizations.

Since its creation, Stroome has been used by USC’s Annenberg School of Journalism and Communication, regional journalism programs including Columbia College in Chicago, and grassroots citizen journalism sites such as

FreeSpeech TV and the Bay Area Video Coalition. Stroome was also used in February 2011 by Egyptian protesters during the Egyptian Revolution when the government shut down social media channels such as Facebook and Twitter. The project initially targeted journalism classrooms and citizen journalism sites as its early adopters, but its user base shifted over time to include more high-‐school educators and students, rather than universities and professional journalists.

By fall 2012, Stroome was working to raise the money to sustain the project and to fix old functions and add new ones to the platform. In an effort to address the project’s reliance on third-‐party vendors which had hampered its ability to fix bugs and build new feature sets for the tool, Stroome eventually brought on a chief technology officer with a focus on rebuilding the platform’s video editing feature. Going forward, Stroome hopes to create a mobile version of the platform, to develop a white-‐label version of the product to be marketed to the corporate community, and eventually to begin charging fees for use.

0

100

200

300

400

500


Social Media Mentions of Stroome

blogs news twiber



TileMill (Tilemapping)


TileMill Development Seed A suite of open-‐source tools that local media can use to make custom, embeddable, hyper-‐local maps

$76,960

TileMill is a project of Development Seed, a data visualization and mapping firm based in Washington, DC. TileMill significantly lowers the barrier to entry for creating highly customized maps, with the aim of allowing journalists and bloggers to tell richer stories and provide unique analysis on local issues through hyper-‐local, data-‐filled maps. THE INNOVATION

TileMill is a free, open-‐source mapping tool for creating highly customized maps, viewable on any web browser and on various mobile devices. Although other basic mapping tools such as Google Maps, OpenLayers, and Polymaps have already made it easier to load a map into a website and plot certain points on it, TileMill offers the ability to change the appearance of base maps (both in design and the data they show) and to easily customize data points. TileMill requires less technical experience than traditional GIS mapping and turns mapmaking into a task that those who are comfortable with common web design languages (HTML and CSS) can quickly grasp. Users customize their maps with the platform’s web-‐based interface and CSS style sheets. TileMill can import maps and layer data from several popular file formats. Maps made in TileMill can be exported and edited through popular software such as Adobe Illustrator. IMPLEMENTATION

Rather than start from scratch, Development Seed built TileMill from a suite of open source libraries including Mapnik, node.js, backbone.js,

express, and CodeMirror.1 Development Seed received funding through the Knight News Challenge to develop TileMill into TileMill 2.0, with a new focus on ease of use. Initial contract negotiations with Knight and discussions on the type of open-‐source license that would be used delayed the project’s launch by a month. TileMill 2.0 was launched on February 16, 2011 with extensive built-‐in help text and thorough documentation available at http://mapbox.com/tilemill/. Development Seed announced this launch through its blog and other social media channels such as Twitter and Delicious. Within the first day of TileMill 2.0’s launch, more than 10,000 people read the announcement on Development Seed’s blog and 1,750 visited TileMill’s website.

Throughout TileMill’s development process, Development Seed regularly asked both the developer community and regular users for their feedback on the tool. Development Seed has released eight updates of the tool since the release of TileMill 2.0, adding a number of key features including a one-‐click installer and a Microsoft Windows compatible version of the software. Version 0.10.0, TileMill’s most recent version, was released in late September 2012. TileMill 0.10.0 offers even more functions for compositing layers and allows for Photoshop-‐like clipping, masking, blurring, and highlighting.

At the start of the project, Development Seed planned to focus its marketing efforts on targeted outreach to journalists and bloggers in the Washington, DC area. TileMill would work closely with the Washington Examiner as a beta 1 Backbone.js is also a component of DocumentCloud, a 2009 Knight News Challenge winner.



partner, to test the software and provide training on how to use the tool to create custom maps for the Examiner’s local stories. TileMill’s point of contact left the Examiner just before TileMill 2.0 was launched, and the project was left to find another beta partner. Despite this initial setback, the tool received interest from a number of mainstream news organizations outside the DC area. TileMill’s team spent more time training these groups in an effort to gain a number of strong samples that could be used to show TileMill’s work and help spread usage to more local bloggers and news organizations. Development Seed also continued promoting TileMill through its blog, social media channels, presentations and trainings at local meet ups and conferences, and through its involvement in the developer community.

REACH AND OUTCOMES

Early signs of impact such as TileMill’s number of downloads, high-‐end clients, and the buzz surrounding the project suggest that TileMill may be well on the way to becoming an essential newsroom tool. It experienced a significant and steady growth of visitors who came to its website for information on the project and/or who downloaded the application to create and modify their own maps. As of early October 2012, TileMill had been downloaded nearly 65,900 times. Among the organizations to download the application are: well-‐known news organizations such as the Chicago Tribune, the Boston Globe, NPR, USA Today, the New York Times, and the Guardian; universities including Cornell; international agencies like Amnesty International; and

0

200

400

600

800

1000


Social Media Mentions

blogs delicious hackernews twicer

0

2,000

4,000

6,000

8,000

10,000

12,000

14,000

16,000

Oct-‐11 Nov-‐11 Dec-‐11 Jan-‐12 Feb-‐12 Mar-‐12 Apr-‐12 May-‐12 Jun-‐12 Jul-‐12 Aug-‐12 Sep-‐12

Pageviews

Pageviews Linear (Pageviews)



government agencies including the White House and the Department of Energy. While TileMill originally planned to target its outreach efforts to the Washington, DC area, the evidence suggests that it has gained considerable awareness and spread to a much broader audience. From January 2011 to October 2012, TileMill was mentioned a total of 10,600 times over social media channels, with 38 percent of those mentions coming from outside the United States. The majority (85 percent) of these mentions were over Twitter. Peaks in TileMill’s mentions in January 2011, September 2011, and January 2012 coincide with release of new versions of the software.

TileMill achieved its broader goal of helping journalists and bloggers tell richer, more complex stories through hyper-‐local, data-‐filled maps. MapBox—which began as Development Seed’s fee-‐based platform for hosting TileMill maps and now exists as a fully separate company that operates TileMill—features a wealth of visually stunning maps that succeed in this goal through the use of the tool.2 TileMill has received a largely positive response from many users, including excitement from Brian Boyer, one of the Chicago Tribune’s main developers of news applications. In addition to the tool’s actual use, project team members are frequently asked to present and give trainings on TileMill at conferences and events around the country.

The project has also attracted significant attention in the area of open-‐source code and cultivated a strong developer community. By late July 2012 TileMill’s code had been downloaded more than 56,000 times, forked 114 times, and attracted nearly 1,000 individuals who signed up to be notified of code changes. TileMill was featured in Linux Magazine as a project to watch, and Development Seed was asked to present on TileMill at a number of open source

2 MapBox’s gallery of web maps built using TileMill is available at http://mapbox.com/tilemill/gallery/.

conferences, including WhereCampPDX, WhereCampDC, POSS4G, and State of the Map.


KNC report

Government & Nonprofit

Transcript of KNC report