“We cannot solve problems by using the same kind of thinking, we used when we created them”
Mark Penn
5
The 5 Ms
Changing the game
6
The 5 Ms Meters
Mobile
Mixed
A new generation of low cost, passive, software-based meters
The next frontier for data collection
Hybrid methods to maximise response and compliance
7
Models
Sophisticated models, ascription and fusion procedures to enhance databases
Management
Co-ordinating and integrating the multiple sources into coherent, usable information
Meters
The next generation
8
Meters: the next generation Passive
Software-based
9
•
SELF INSTALLED
•
SIMPLE PROCEDURE
•
MINIMAL MAINTENANCE
•
RESPONDENT-FRIENDLY
•
NO NEED TO REMEMBER WHAT IS WATCHED OR LISTENED TO
•
JUST TO KEEP IT SWITCHED ON, CHARGED AND CLOSE
•
AS PEOPLE NORMALLY DO…
Low cost •
LOW COST PER PANELLIST
•
MORE FLEXIBLE CONTRACTS
•
LARGER PANELS
Ipsos innovation: • The most advanced passive, electronic measurement on the market • Data capture via Smartphone detection of encoded signals • No button pressing or extra devices to carry • Developed and tested in multiple environments
10
TV
11
ALL PLATFORMS
ALL VIEWING
Television sets
Live
Tablets
Time-shifted
Smartphones
On-demand
ALL THE TIME
ALL LOCATIONS
Out of home viewing captured by Smartphone
In-home Out of Home
Mobile
Always On…
12
Mobile: always on…
Growing coverage
13
Growing usage
Always on
Mobile research: example • Multi-media audience study (UK, USA and Canada) • Integrates separate audience measurement studies via a ‘hub’ survey • People asked to indicate media usage and moods each half hour on their Smartphone
14
Mixed methods Reaching everyone
15
Mixed methods: reaching everyone
Not everybody answers the door 16
Few answer their landlines – if they have them
Not everybody is online…
So we use mixed methods to reach them Sample recruitment
Data collection 62% 12% 16%
91% CATI/ 9% F-t-F 17
10% Face-to-Face
90% Online
Models
Adding value to surveys
18
Models: adding value to surveys Data fusion
Data ascription
Data modelling
• SOME QUESTIONS MAY BE MISSED BY RESPONDENTS
•
SOME SURVEYS ARE TOO LONG FOR ONE PERSON TO ANSWER
•
• TO OBTAIN A COMPLETE RECORD, WE CAN ASCRIBE ANSWERS GIVEN BY SIMILAR PEOPLE
•
WE CAN FIELD SEVERAL SHORTER SURVEYS, ASKING A CORE SET OF QUESTIONS TO ALL RESPONDENTS
CERTAIN DATA CANNOT BE AFFORDABLY COLLECTED IN A SURVEY
•
IN SOME CASES, IT IS POSSIBLE TO COMBINE DATA WE DO HAVE WITH LIMITED SURVEY DATA TO PAINT AN ACCURATE PICTURE OF RESPONDENT BEHAVIOUR
19
•
…AND DIFFERENT QUESTIONS TO MATCHED SETS OF PEOPLE
•
WE THEN ‘TRANSFER’ DATA FROM MATCHED PEOPLE BETWEEN SURVEYS
Models: examples Data ascription
TWO PRODUCT CONSUMPTION SURVEYS RUN FOLLOWING OUR MAIN MEDIA STUDY. DATA FROM SURVEY 1 ASCRIBED TO RESPONDENTS FROM SURVEY 2 (AND VICE VERSA) AND INTEGRATED INTO MAIN STUDY RESULTS
20
Data fusion
COMSCORE PANEL DATA FUSED WITH SURVEY DATA FROM OFFICIAL BRAZILIAN READERSHIP STUDY TO PROVIDE CROSS-PLATFORM AUDIENCE INFORMATION FOR NEWSPAPERS AND MAGAZINES
Data modelling
LOCAL NEWSPAPER READERSHIP MODELLED USING COMBINATION OF CIRCULATION, POPULATION DATA AND SURVEY INPUT
Management
Pulling it all together
21
Managing change, managing complexity •
Surveys are rarely sufficient to provide a comprehensive overview of audiences now
•
We need to measure cross-platform audiences, audience engagement, product consumption, audiences to smaller content distributors and more
•
We need to offer analytical software and data visualisation
•
It must be Better, Faster and Cheaper than before
•
We have the technologies and the methodologies to do this
•
But we need to navigate the changes this will cause in long-standing audience ‘currencies’
•
At the end, this will be about skilled people and openness to change