Aka

61
5 th February 2014

description

 

Transcript of Aka

Page 1: Aka

5th February 2014

Page 2: Aka

This morning…

1.Intro’s etc

2.Lifting the lid (on your data)

3.DataShaka; go on then, impress us…

4.Further questions, next steps etc

Page 3: Aka

Intro’s…

Page 4: Aka

The Data Journey…

Page 5: Aka
Page 6: Aka

The agency problem…

Page 7: Aka

Client A

A/C Team

Solution

Vendor

Client B

A/C Team

Solution

Vendor

Client D

A/C

Team

Solution

Vendor

Client C

A/C Team

Solution

Vendor

Client E

A/C

Team

Solution

Vendor

Page 8: Aka

Client A

A/C Team

Solution

Client B

A/C Team

Solution

Page 9: Aka

Client A

A/C Team

Solution

Client B

A/C Team

Solution

Page 10: Aka

Client A

A/C Team

Solution

Client B

A/C Team

Page 11: Aka

Client A

A/C Team

Solution

Client B

A/C Team

Client C

A/C Team

Solution

Page 12: Aka

Client A

A/C Team

Solution

Client B

A/C Team

Client C

A/C Team

Solution

Client D

A/C Team

Solution

Client E

A/C Team

Page 13: Aka

Client A

A/C Team

Solution

Client B

A/C Team

Client C

A/C Team

Solution

InternalClient A

IT

Solution

Internal Client B

Finance

Solution

Page 14: Aka

The brief.

Page 15: Aka

1. Ticketing - Database +

Analysis Tool

2. More sources

3. Future needs…

Page 16: Aka

The Ticketing Journey…

Page 17: Aka

AudienceAgents Theatres

ENTAAudience View

Ticket MasterTessitura

Clients

The Brief.

Page 18: Aka

The Brief. 1. Creating a Database

1. Creating an analysis tool

2. More Sources

Page 19: Aka

3. Future needs

Evolving SourcesEvolving Query Space

Evolving AKA User NeedsEvolving Client Needs

Page 20: Aka

Lifting the lid…

Page 21: Aka

A small sample came with the brief

A bigger example set requested

LearningsAction

1. Multiple Formats for ‘the same’ conceptual set

2. Different formats from the same system

3. Data quality/format issues1. Ticket data all delivered by email2. Data in CSV, XLS and PDF3. A lot of manual work4. Historical data stored in folders in Excel5. Local server ‘almost full’

Examining what we received

1. MSG files can be automatically extracted

2. Data can be automatically extracted from PDF

3. All three formats must be handled4. Email must be handled5. There is duplication in the files

Core domain data set exploration 1. Waste within data acquisition

Page 22: Aka

Duplication and Waste: an efficiency opportunity…

• 30th December 2013 to 26th January 2014

(28 days)

• 4 different report types split overo Advanceo Dailyo Wrap

• CSV and PDF file format

• 115 files

• 47.6mb

• Most detailed level:

o Price Type

• Waste due to pivots and aggregates:

approx. 25%

o 3 report areas, 4 reports = 1

unnecessary

Page 23: Aka

Duplication and Waste: an efficiency opportunity…

• 30th December 2013 to 26th January 2014

(28 days)

• 4 different report types split overo Advanceo Dailyo Wrap

• CSV and PDF file format

• 115 files

• 47.6mb

• Most detailed level:

o Price Type

• Waste due to pivots and aggregates:

approx. 25%

o 3 report areas, 4 reports = 1

unnecessary

• 30th December 2013 to 27th January 2014 (29

days)

• 12 different report types split overo Advanceo Maturedo On-Day sales

• XLS and PDF file format

• 256 files

• 45.7mb

• Most detailed level: o Seat Type by Price Band by Discount

type by Performance Type

• Waste due to pivots and aggregates: approx.

75%

o 3 report areas, 12 reports = 9

unnecessary

Page 24: Aka

LearningsAction

1. Multiple Formats for ‘the same’ conceptual set

2. Different formats from the same system

3. Data quality/format issues1. Ticket data all delivered by email2. Data in CSV, XLS and PDF3. A lot of manual work4. Historical data stored in folders in Excel5. Local server ‘almost full’

1. MSG files can be automatically extracted

2. Data can be automatically extracted from PDF

3. All three formats must be handled4. Email must be handled5. There is duplication in the files1. Waste within data acquisition2. A rich and valuable core set3. Many connection points for other sets

A small sample came with the brief

A bigger example set requested

Examining what we received

Core domain data set exploration

Page 25: Aka

Report Date

Unpaid

Count

Unpaid

Net

UnpaidCharges

Unpaid

Total

Seats Comps Value Capacity TargetReserved

SeatsReserved

Value…*Potential

Value

Source

Show Theatre

ShowType

ShowMonth

ShowYear

ShowDay

ShowDOW

Discount Type Price

BandSeat Type

PaymentChannel Agent

CardType

Page 26: Aka

Show/Theatre

Location/Theatre

Segment

Ticketing

PaidMedia Social

Media

Email

Print Media

BroadcastMedia

Weather

Events

Out of homeMedia

Page 27: Aka

And we only looked at 2 shows over 30 days…

Page 28: Aka

Analysis Tool

Database

Page 29: Aka

DataShaka…

Page 30: Aka

Store

Harvest

“Everything is a source...”

http

file

FTP

email

API

market

place

secure

server

Unify

DeliverTimeT

UnifiedData

Context

C

SignalS

ValueV

Page 31: Aka
Page 32: Aka

ConsilientConnection

istChaordicRobustFlexible

Any question of any data.

32

Page 33: Aka

Time Context Signal Value

Context Type

Ct

Page 34: Aka

Sales

Ct

34

Page 35: Aka

{ "name":"Nutrigum" "followers_count": 39061, "friends_count": 12986, "listed_count": 917,}

Harvested at 2013-10-01 17:35:00

{ "category": "Company", "talking_about_count": 58550, "username": "healthyx", "likes": 1985655, "link": "http://healthyx"}

Harvested at 2013-10-01 19:12:00

<performance> <account> Healthy X Limited </account> <cam>nutrigum – branding</cam> <data> <date v=“2013-10-01”> <impr>14000</impr> <clk>1500</clk> <cnv>10</cnv> </date> </data></ performance >

Adserver

Sales <filterTags>Nutrigum</filterTags><tagStats> <tag>~SOURCE~t</tag> <tagDisplayName> TWITTER </tagDisplayName> <matchCount>71</matchCount> <popularity> <popularityCount> <timeInterval> 2013-10-01 </timeInterval> <count>2.0</count> <normalizedCount> 2.0</normalizedCount> </popularityCount> </popularity> </tagStats>

eCRMDate,productId,userId,number,ppu2013-10-01,123,321,2,5.002013-10-01,123,521,1,5.002013-10-01,333,444,2,15.002013-10-01,854,111,1,20.00

Some Data…

Page 36: Aka

Store

Harvest

“Everything is a source...”

http

file

FTP

email

API

market

place

secure

server

Unify

DISQ

Unstructured

Relational

Graph

In Memory

Document Store

File System

Big Table

Deliver

Enterprise Data Store

TimeT

UnifiedData

Context

C

SignalS

ValueV

Page 37: Aka

Your data in TCSV…

Page 38: Aka
Page 39: Aka

Lots of TCSV

Page 40: Aka

Viewing & analysing……

Page 41: Aka
Page 42: Aka
Page 43: Aka
Page 44: Aka

Other ways to view…

Page 45: Aka
Page 46: Aka
Page 47: Aka
Page 48: Aka

Clients

Page 49: Aka
Page 50: Aka

Ways of working

Page 51: Aka

Partnershi

p

Page 52: Aka

An Agile Approach

Lean

Change

Collaboration Communication

Agile

Page 53: Aka

Pitch

Statement of Work signed

off

Follow-up:Tech Deep Dive

SurgeryWorkshop

1.Analysis Tool

1.TicketingDatabase

2.More

Sources

3.The Future

6-8 weeks 8-10 weeks

Page 54: Aka

How Much?

Page 55: Aka

Phase Configure Monthly Timing1. Ticketing £25k £5k 6-8 weeks

2. More sources £15k Incl. 8-10 weeks

3. Future TBC TBC TBC

Total £40k £5k

Page 56: Aka

There is not one answer.

Page 57: Aka

Solutionapproach Pro’s Con’s

Re-Seller

Custom Build

Data As A Service

Page 58: Aka

Solutionapproach Pro’s Con’s

Re-Seller• Off the shelf ‘modules’• Polished• Safe buy

• Lack of influence• Take ‘as is’• ‘just’ a sales team

Custom Build

Data As A Service

Page 59: Aka

Solutionapproach Pro’s Con’s

Re-Seller• Off the shelf ‘modules’• Polished• Safe buy

• Lack of influence• Take ‘as is’• ‘just’ a sales team

Custom Build• Get what you want• Direct involvement• Can be cheaper

• Get what ‘only you’ want• Focussed on ‘now’• Or more expensive

Data As A Service

Page 60: Aka

Solutionapproach Pro’s Con’s

Re-Seller• Off the shelf ‘modules’• Polished• Safe buy

• Lack of influence• Take ‘as is’• ‘just’ a sales team

Custom Build• Get what you want• Direct involvement• Can be cheaper

• Get what ‘only you’ want• Focussed on ‘now’• Or more expensive

Data As A Service• Proven specialist platform• Fully configurable• Future proof

• Data vs DV specialists• Requests not instructions• Active involvement

Page 61: Aka