Measuring online campaign success

A to Z approach

Global Inputs H1 2009 Internet ad spend is growing despite the credit crunch: in UK (23.5%)

it beats TV (21.9%), and grows 4.6% in 2009, while total ad spend goes down massive 16.6%.

Internet measurement becoming as advanced as ever, joining together site-centric, campaign-centric and video measurement tools.

CEE markets catch up with broadband penetration levels and inherit best practices on serving heavy ad content

Current Internet development allows brands to interact with their users more and more, engage them into social, community, creative, fun and gaming activities – all the innovative things which you can think of, when planning communication strategies

Local inputs In Lithuania internet penetration is 54% (7 y.o. +), which equals 1.7 mln users Online adspends share – expected around 5.5% this year, same as last year

due to crisis influence. Online ad market volume expectation - 22 mln lt in 2009.

On average Lithuanian internet users spend 12H online monthly, and 1H average on a regular working day – that is 2-3 times below European averages

Broadband takes 42% of household connections Thanks to it – all major innovative tools such as video, microblogging, viral

games are there.

Internet media is already huge. So how to make most of it? What does it mean - to measure campaign success and user

engagement from A to Z?

Imagine a case-study…You are launching a new product with Your client

You have clear idea about your TG, offer characteristics, pricing etc via regular consumer intelligence research

You have selected Your creative agency, web-developers

You have developed a product website with hot & trendy web 2.0 elements such as games, tests, social networking, video content

Now it is the time to launch the communication strategy…

Pre test

Basic ad serving data

CTR

Conver-sions

Actions

Influence on web site

Audience profile

Video/ Buzz

Behavio-ral tracking

Branding test

Tag Your site Test Your ads

Serve the CampaignGet users to Your site

Profile themTrack themAsk themFollow up

Single-source solution from concept creation to evaluation

1 - Start with online copy-test

Set a TG, matching predefined socio-demographic criteria Determine what character / content is the most clear and attractive Examine which creative is most associated with the advertised brand

and TG Investigate each creative power to encourage a visit to advertised

web site / to take part in a contest / to make a purchase

Exemplary test: content features

QUESTION: Please mark to what extent you agree or disagree with the following statements concerning this advertisement.

ad 1: N = 500ad 2: N = 500ad 3: N = 500ad 4: N = 500

• ad 1 seems to be the leader in regard to attractiveness and drawing attention

• ad 2 seems to be the leader in regard to clarity

• ad 4 seems to be the worst evaluated creative in regard to all the dimensions

the chart presents the average levels of agreement with the statements (on the scale 1 means “I definitely disagree” and 7 means “I definitely agree”); answers “I don’t know” were excluded from the calculation

1,0

2,0

3,0

4,0

5,0

6,0

7,0It is attractive

It has 'that something'

I like it

It is legible

It is clearIt makes sense

It draws attention

It is distinctive

It is noticeable

ad 1 ad 2 ad 3 ad 4

„Attention drawing”

„Clarity”

„Attractiveness”

QUESTION: What would you associate this advertisement with at first glance?

ad 1: N = 500ad 2: N = 500ad 3: N = 500ad 4: N = 500

the arrows on the chart are always directed towards the group with the higher percentage of indications of a given association than the other group

the sum of the percents is not equal to 100 as respondents could have given more than one answer

• most common associations with all the creatives are beauty and care products and an exotic trip

• the following creatives were relatively most often associated with the particular content: ad 2 and ad 4 – with a contest ad 2 – with beauty and care products ad 1 – with an exotic trip ad 4 – with an party

47%

50%

5%

5%

2%

6%

4%

59%

45%

5%

12%

1%

4%

1%

46%

31%

8%

4%

9%

8%

7%

39%

40%

16%

11%

5%

6%

4%

0% 10% 20% 30% 40% 50% 60% 70%

an advertisement ofbeauty and care products

an advertisement ofan exotic trip

an advertisement ofa party

an advertisement ofa contest

an advertisement ofa club

something else

I don't know/hard to say

percentage of respondents

ad 1 ad 2 ad 3 ad 4

Exemplary test: content associations

Creative type CTR-U CR-U Effective CTR

Leaderboard 0,57% 79,17% 0,45%

Full Banner 0,65% 81,31% 0,53%

Skyscraper 0,70% 77,13% 0,54%

Square Button 0,91% 81,56% 0,74%

iLayer 1,87% 62,93% 1,18%

Interstitial 3,70% 51,69% 1,91%

2 – Serve and audit ad campaign

From one side – clicks are important, but is it ALL we can take from campaign?

Czech Republic stats, 2008

So what does CTR really mean?

CTR

CR

Post-Click

Post-View

USERS seen creative

USERS clicked

USERS moved to target page after clicking

USERS performed action on target page after

clicking

Going deeper with direct campaign response

USERS moved to target page after seeing creative

Joining data from adserver and web analytics

ALL visitors from your campaign

may actively use your web-site

From other side – not-clicking visitors always dominate, do not underestimate their importance…

3 – Track users’ actions

! DO they follow the same behavioral patterns as regular visitors?! DO they make required actions?! DO post-view visitors do what you want them to?! DO they become loyal users and “core audience” of advertised web-site?

What happens next with people redirected to target web-site?

Example of visitors behaviour

What might have gone wrong? Maybe You simply attracted not the right people?

Following is some real example…

Let’s say: You run a campaign of a well-known car brand, targeted for persons aged 25-39, with secondary or higher education, living in medium and small cities …

…You get your average CTR of ~1%, conversion rate at ~80%

…Within post-campaign analysis You discover that some required actions were made, but loyal audience (especially post-click) grew insignificantly…

4 – Profile your audience

Car brand campaign CTR profile

With innovative tracking solutions such profiling is possible “on-the-fly”:

! NO need to wait 1-2 weeks after campaign end to produce the results

! TRACK now – measure effect now.

! ONLY this will allow to actually influence and change things before its too late.

!AND that is what Internet is for = instant reaction to events = saving money

Instant XLS report –

views, clicks, actions

Full Campaign/Site AudienceVisitors

(cookies)Audience

composition Page views

Average number of page

views per visitor

Average time per

visitor

Age:18-24 1,651,653 23% 3,446,105 2.09 0h 0m 10sAge:25-34 2,867,304 40% 5,562,277 1.94 0h 0m 10sAge:35-44 1,447,905 20% 2,876,022 1.99 0h 0m 12sAge:45-54 811,875 11% 1,582,627 1.95 0h 0m 8s

Age:55+ 439,494 6% 836,717 1.90 0h 0m 6s

Euroset_SamsungS5230 |impressions|

Yandex.ru-main

4 – Profile visitors’ behavior

Why post-branding tests matter?

• FMCG, Auto-Moto and other traditional companies are stepping stronger into Internet. Telecoms are long living there = their main concern is brand promotion.

• Internet changes to main support medium for TV = Advertisers expect same capabilities and even more.

Online branding studies (pre & post) in form of on-site surveys help to evaluate user brand experience.

5 – Evaluate influence on brand

QUESTION: Please list up to 5 brand names of ANALYSED TYPE OF PRODUCT that first come to your mind.

100% - all the respondents

Pretest: N = 1 000Posttest: N = 1 000

Posttest (mailing): N = 680

• BRAND X is the unchallenged leader in regard to the top of mind awareness in the general sample

• there were no significant differences in BRAND X results between the groups that filled in the questionnaire before the campaign (pretest) or after it (posttest)

67%

9%

9%

4%

3%

2%

2%

1%

3%

0,3%

66%

9%

7%

8%

3%

2%

2%

1%

3%

0,4%

83%

4%

3%

3%

2%

2%

0,4%

0,3%

1%

0%

0% 10% 20% 30% 40% 50% 60% 70% 80% 90%

Brand X

B1

B2

B3

B4

B5

B6

B7

other

no answer

percentage of respondents

Pretest

Posttest

Posttest (mailing)

Branding studies – spontaneous awareness

0%

10%

20%

30%

40%

50%

60%1

2

3

4

5

6

7

8

9

10

Pretest Posttest Posttest (mailing)

18

QUESTION: We would like to know your opinion concerning particular brands of ANALYSED TYPE OF PRODUCT. Please check all the brands that fit the following statements.

100% - all the respondents who know BRAND X

Among the general sample BRAND X is significantly better evaluated by those who filled in the questionnaire before the advertisement campaign in respect to criterion 1 (products of this brand perform longer than others)

1 - products of this brand perform longer than others

2 - This brand is close to me

3 - This brand has High-Tech products

4 - products of this brand have higher quality than others

5 - This brand is reliable and trustworthy

6 - products of this brand offer good quality-price relation

7 - It's an expert in the product type 1 sector

8 - It's an expert in the rechargeable product type 1 / type 2 branch

9 - These are best products to supply portable music players

10 - It's the best solution for High-Tech devices Pretest: N = 962

Posttest: N = 945Posttest (mailing): N = 671

Branding studies – brand equity

Mounting an effective campaign in online always means integration with research functions.

Pre-research, monitoring, and post-analysis used standalone do not provide full picture – solution should be integrated: one source data highly advised.

Digital marketing is a tool - one of the best ever created, considering the perspective possibilities.

But doing it right ain't cheap, or easy .

The bottom line is…