CRE QUARTERLY MEETINGDECEMBER 3, 2015
OPENING COMMENTS, CERIL SHAGRIN
OPERATING ASSUMPTIONS
NIELSEN R&D UPDATE
CHRISTINE PIERCE
DECEMBER 2015
CRE QUARTERLY RESEARCH UPDATE
Copy
right
©20
13 T
he N
ielse
n Co
mpa
ny. C
onfid
entia
l and
pro
prie
tary
.
6
RESEARCH UPDATE
• Update on Program Names Research (Follow up from September Meeting)
• Nielsen Feedback from Nielsen/CRE/GfK session
• Reflections on 2015 & Looking ahead to 2016• How can the CRE help Data Science be more effective?• How can Data Science help the CRE be more effective?
STEERING COMMITTEE
PAT LIGUORI
Membership Steering Committee voted unanimously to recommend the below individuals for full CRE membership. Both have actively participated on CRE committees.
• Mainak Mazumdar, Chief Science Officer, Simulmedia˃ Recommendation: engage with a CRE committee
• Ann Casey, Corporate Research Director, Weigel Broadcasting˃ Prior service on CRE Committee; awaiting bio and statement
Seeking Committee on which to Participate• Bernie Shimkus, VP Dir of Media Research & Consumer Insights, Harmelin
Media
Fourth Quarter 2015 Update
Steering Committee
RESEARCH COMMITTEE REPORTS
DATA QUALITY
CERIL SHAGRIN
LOCAL MEASUREMENT
BILLY MCDOWELL
Billy McDowell Raycom MediaChair
Andy Rainey RABAnn Casey WeigelBuzz Knight Greater MediaBruce Hoynoski NielsenCeril Shagrin UnivisionDave Daniels ABCHadassa Gerber TVBJoanne Burns FOX TelevisionJohn McMorrow CoxReps
Janice Finkel-Greene Magna GlobalVice Chair
Julie Russell Adco AgencyKathleen Bohan UnivisionKeenan Pendergrass WFTV OrlandoLucy Hughes Media GeneralMark Kaline Kaline MediaPat Ligouri ABC TV StationsRick Pike IntermediaSarah Smith HearstScott Osborne UnivisionTim Daly ITNTony Marinaro Media General
Richard Zackon Facilitator
Committee Members
Local Measurement Report
The project was a follow up to the study last fall and used predictive modeling to estimate local market TV ratings with the following additional information to inform the models:
1. Program Names2. Return Path Data from Set Top Boxes
During the Fall 2015 Quarterly meeting, the council approved $335,000 to fund a predictive modeling project in conjunction with Nielsen and Columbia University. Richard Zackon served as project leader.
The project was terminated by the committee in November 2015 due to the lack of availability of return path data from Nielsen.
Local Measurement Report
This project was the third project from the CRE which planned to use return path data but that data was not provided by Nielsen.
1. Set Top Box Data Committee2. Sample Quality Committee3. Local Measurement Committee
The committee also expressed reservations about Nielsen’s ability to operationalize return path data as Charter is the only source Nielsen currently accesses.
Local Measurement Report
The committee will revisit the project in early 2015 and look at other sources of available data including Smart TV data.
All project members will be paid according to their time spent in 2015.
DIGITAL RESEARCH
BRAD ADGATE
BIG DATA
STACEY SCHULMAN
Validation of Audience Attributes & BehaviorsRESEARCH PROJECT FUNDED!
Real World Performance• Empirical Research with live, in-market test, to determine ability of a
campaign to deliver true target audiences and the variations across Big Data vendors
• Requires a “truth set” for comparison (known currencies or original fieldwork)
Understanding Big Data Consumer Targeting Techniques• Educational exploration to understand variations in marketplace
approaches to consumer target development and ad delivery• Includes Data Enrichment Providers (DEPs), 3rd Party Data and Ad
Tech firms• All digital environments considered• Goal is uncover factors that impact the degree to which Big Data
vendors can accurately identify consumer targets, and ultimately work toward a set of minimal standards for the marketplace
Timing• RFP posted in early September. Project awarded to Pre-Meditated
Media on November 24th (thank you!) Delivery early Spring.
Validation of Audience Attributes & BehaviorsRESEARCH PROJECT FUNDED!
Critical Data Quality Issues to Explore (a starting point)
• Recency - The impact of recency of data collection on validity of both past and predictive behavior
• Stationary v. Modality – How does the rate at which an attribute or behavior change state impact accuracy? (gender, employment, marital status, age, salary, product purchase intent, etc)
• Audience Representation – How can we uncover data blind spots
• Predictive reliability of audience characteristics & behaviors. Are certain characteristics better predictors of behavior?
Validation of Audience Attributes & BehaviorsRESEARCH PROJECT FUNDED!
Research Design – 2 Phases
• Phase 1 - Marketplace Data Validation Interviews (8 DEP firms) to explore:
• Modeling process procedures• Validation of modeled targets in generating consumer response• Validation of modeled targets in delivering the true target audience in terms of audience
composition• Access to and use of “truth” data sets for authentication of true target audience• Source diversity of data, how many different data sets used to formulate models• Source history of data; ie., some data could be carried forward as part of a data mash up• Recency or freshness of data• Representativeness of data sets, including demography and geo location
• Phase 2 – In-Market Campaign Test (8 DEP firms)• Controlled test to determine the ability of a live campaign to deliver true target audiences
and how this delivery might vary across DEPs and advertiser categories.
Validation of Audience Attributes & BehaviorsRESEARCH PROJECT FUNDED!
Research Design – 2 Phases
• Phase 2 – In-Market Campaign Test (8 DEP firms)• DEP / Category Intensive – Tests modeled target efficacy across eight DEPs across four advertiser
categories. Highlights the strengths and limitations of certain DEPs to accurately identify target audiences.
Matching Modeled Targeted Ad Impressions to a CRM Truth File
Scenario 3Four Categories x 8 DEPsEight DEPs per Category
Data Enrichment Provider A B C D E F G HAd Impressions 500,000 500,000 500,000 500,000 500,000 500,000 500,000 500,000 Truth File Match (Uniques) 22,699 32,568 24,555 35,668 10,997 12,456 18,766 13,478 Truth File Match Rate 5% 7% 5% 7% 2% 2% 4% 3%
Auto Intender Tgt IncidenceModeled Target Placement
Data Enrichment Provider A B C D E F G HAd Impressions 500,000 500,000 500,000 500,000 500,000 500,000 500,000 500,000 Truth File Match (Uniques) 32,255 43,568 22,146 73,547 43,778 54,211 23,456 64,776 Truth File Match Rate 6% 9% 4% 15% 9% 11% 5% 13%
Modeled Target PlacementInsurance Intender Tgt Incidence
Data Enrichment Provider A B C D E F G HAd Impressions 500,000 500,000 500,000 500,000 500,000 500,000 500,000 500,000 Truth File Match (Uniques) 67,543 82,567 72,887 91,223 69,300 47,156 58,990 61,224 Truth File Match Rate 14% 17% 15% 18% 14% 9% 12% 12%
Kitchen Appliance Tgt IncidenceModeled Target Placement
Data Enrichment Provider E F G H I J K LAd Impressions 500,000 500,000 500,000 500,000 500,000 500,000 500,000 500,000 Truth File Match (Uniques) 76,555 44,212 89,012 93,002 67,890 55,780 59,877 59,653 Truth File Match Rate 15% 9% 18% 19% 14% 11% 12% 12%
Modeled Target PlacementSmart TV Tgt Incidence
Validation of Audience Attributes & BehaviorsRESEARCH PROJECT FUNDED!
Truth File Creation
To be supplied by a DEP with substantial CRM data base coverage of U.S. consumers and the email address linkage to map back to cookie pools.
Advertiser Recruitment• Advertisers will be recruited from moderate to strong data-intensive
categories (automotive, finance, insurance) or high-incidence consumer categories (home electronics).
• Advertisers will be sourced using a combination of CRE and Pre-Meditated Media contacts.
• Since CRE is underwriting media costs, advertisers will be receiving media weight incremental to their current in-market media plans.
Validation of Audience Attributes & BehaviorsRESEARCH PROJECT FUNDED!
Timeline
Validation of Audience Attributes & BehaviorsRESEARCH PROJECT FUNDED!
Costs & Considerations• $240K inclusive of media costs
• Considerations:- Confidentiality
EVALUATION OF INDIVIDUAL PLATFORMSSUMMARY OF FINDINGS
JOE ABRUZZO
Major issues raised during expert interviews* included:− Definition of “cross-platform” - what to include− Differences in the definition of an exposure, an impression− Measurement deficiencies for individual platforms− Valuation of individual platform impacts and estimation of total media plan
performance (how to combine impressions, GRPs, etc.)− The impact of concurrent cross-platform usage
* Interviews were conducted in the fall of 2014 by Horowitz Associates and Betsy Frank Insights
THE ORIGIN: INTERVIEWS CONDUCTED AMONG INDUSTRY EXPERTS
How does viewing on a TV set, computer, tablet or smartphone differ when it comes to each of the following?
A COMPARISON OF VIEWING EXPERIENCE ACROSS PLATFORMS
ENGAGEMENT Do consumers connect more or less deeply with what they’re watching based on the device they’re using?
IMPACT OF SETTING
Does the viewing situation have a greater effect on engagement for certain devices?
RETENTION AND RECALL
Do consumers retain information about what they’re watching better on some devices than others?
EMOTIONAL REACTION
Does the device used influence feelings toward either show content or the ads they see?
Few differences by platform when it comes to…• Specific emotional reactions to the show viewed• Ability to recall show plot points
Viewers to the TV platform• Rated their level of attention higher than other viewers…• Were more satisfied with the general experience
Those who watched on a phone were the most likely to say they enjoyed the show itself
HIGH-LEVEL SUMMARY OF FINDINGS: THE SHOW
Viewers to the TV platform consistently more likely to…• Say they paid attention to ads• Recall brands airing commercials during the show• Correctly identify specific creative elements in the ads• Correctly match specific ads with specific brands
HIGH-LEVEL SUMMARY OF FINDINGS: THE ADVERTISING
Webinar• 3PM – Thursday, December 17• To be limited to current CRE and committee members
To provide a more in-depth discussion of study findings• Attention, Engagement, and Recall of program content and ads by Platform• Influence of platform vs. other factors
• Mood: present, fatigued, stressed, distracted . . . • Multitasking by type: analog, digital . . . • Presence of others and whether co-viewing• Environment: location, room, position
• Opportunity to ask questions about the research
IMMEDIATE NEXT STEPS FOR THIS RESEARCH
CONCURRENT PLATFORM USAGE
JANET GALLENT
CPU PROJECT UPDATE
UX In-Home and Remote Interviews near completion
Quantitative phases have been completed• Quantitative self-report survey (N=3,165 Gen Pop
13+; and N=1,066 supplemental of Hispanic Pop 13+ for a total N=4,020)
• Passive digital metering of 2 devices in the HH (N=310)
CONCURRENT MEDIA USAGE
US Gen Pop 13+: Total Day• 68% reported at least one concurrent event in an average day
(about 175 M)• Skews Female, HHs with kids, Affluent
US Gen Pop 13+: Primetime• 38% reported at least one concurrent event during primetime in an
average day (about 97 M)• Skews Younger, HHs with kids, Higher Income, Have technology
US Hispanic Pop 13+: Total Day• 68% reported at least one concurrent event in an average day (about
27M)• Skews Younger, Some college, Have technology
TIME SPENT USING MEDIA: AVERAGE DAY
US Gen Pop 13+: Total Day• 11 hours and 58 minutes spent using media• 1 hour and 40 minutes spent using media concurrently • 14% of total day media time is concurrent use with other media
US Gen Pop 13+: Primetime (7 pm – 11 pm)• 2 hours and 34 minutes spent using media during primetime• 21 minutes spent using media concurrently during primetime• 14% of total primetime media time is concurrent use with other
media
US Hispanic Pop 13+: Total Day• 14 hours and 31 minutes spent using media• 2 hours and 16 minutes spent using media concurrently• 16% of total day media time is concurrent use with other media
TOP DEVICES USED CONCURRENTLY: US GEN POP 13+ TOTAL DAY
19 minutes are spent using the TV set & smartphone concurrently (TV set & Tablet: 9 minutes)
23 minutes are spent using the TV set & computer concurrently
13 minutes are spent using the computer & smartphoneconcurrently
8 minutes are spent using audio device & smartphoneconcurrently
In an average day…
TOP DEVICES USED CONCURRENTLY: US GEN POP 13+ PRIMETIME HOURS (7 PM – 11 PM)
6 minutes are spent using the TV set & smartphone concurrently)
6 minutes are spent using the TV set & computer concurrently
In an average day during primetime…
3 minutes are spent using the TV set & tablet concurrently
2 minutes are spent using computer & smartphoneconcurrently
BREAK
CRE AUDIO COMMITTEEDUAL DIARY MEASUREMENT STUDY
Presented byBuzz KnightVP of Program DevelopmentGreater Media, Inc
REVIEWObjective: Evolve Nielsen Audio service in Small & Medium markets to better meet needs of local broadcasters- Provide radio clients localized shopping and purchase information - Highlight local radio’s strength of community and listener engagement- Leverage separate radio, TV and qualitative samples for greater efficiency
Markets:- Bakersfield, CA (Hispanic DST)- Charlottesville, VA (Black DST)
Same set of respondents for TV, Radio & qualitative surveys
TV: personal diary instead of set-based diary. Redesigned to be similar to Radio
Radio: remains traditional diary
Field surveys one month apart- 50/50 split:
• TV first/Radio second • Radio first/TV second
TV DIARY ENTRIES
CRE SMALL MARKETS PROJECTKey Takeaways:1) Matched Sample (Called for Placement)
- Consent performance better than currency—0.5 pts higher for Radio and 2.8 pts higher for TV
- Key to success was using an experience call center vendor when we couldn’t use our internal call center
2) Unmatched Sample (Direct Mailed)- Screener returns a bit lower than production due to shorter field period (due to
very tight timeline of test)- Would expect rates to be fine in production with a normal schedule
3) Diary Returns comparable or higher than production--1456 for 1st Radio Week (nearing completion), 1129 for 1st TV Week(still getting a significant number of returns)—NOTE: Diaries have not been checked for in-tab status yet.
- Will start getting returns from the 2nd Week (Radio then TV group) this week- Will get return from 2nd Week (TV then Radio group) in two weeks- Results from the 2nd week will be important metric—Should have a good read on
final return rates in January
CRE SMALL MARKETS PROJECTKey Takeaways Continued:4) Scarborough Test
- Booklets sent to FA15 Phase 1 radio diary keepers—41% return rate (still getting returns)
- Reminder (with replacement booklet) goes out soon. Still targeting a final return rate of 50-55%.
5) TV Diary- Person-level TV diaries look very good. People understand how to
complete them and what type of viewing should be recorded- Most entries are tagged as “Live” viewing. Less than 5% of entries are
“DVR-Time Shifted.”- Significant amount of OTT (Over the Top) and Streaming media being
reported—particularly for younger diary keepers- Sent scanned copies of two diaries—one is a person who viewed Live
TV, OTT and Streaming services and the other shows a person who reported Time Shifted viewing
ROI
DAVE POLTRACK
SOCIAL MEDIA
BETH ROCKWOOD
NEUROMETRICS
HOWARD SHIMMEL
12-24 VIEWING BEHAVIOR
TANYA GILES
OTHER COMMITTEE REPORTS
INSIGHTS TO PRACTICE
COMMUNICATIONS COMMITTEE
JOANNE BURNS
COMMUNICATIONS COMMITTEE
Nielsen 360 Follow-Up • Propose to Nielsen reviewing results of recent CRE studies at the next
360 Conference.
CRE Social Media • After careful review, we will cease CRE social media representation.
Activity Report
Review Process of CRE Presentations • Seeking two volunteers per project, initially from among low-participant
council members. Shelley will keep track of volunteers. • PR will continue to review grammar and spelling.
New Graphic Artist for CRE Presentations• Close to deciding on new graphic company with a possible second
company as back up.
COMMUNICATIONS COMMITTEE REPORTPublicity Report
Recent & Forthcoming Press Activity
• Meeting at Wall Street Journal, R. Zackon and Nathalie Tadena, media industry reporter
EDUCATION COMMITTEE
JED MEYERS
Activity update
Met with Monica and Meijing (NYU students) to evaluate content they informational / educational content that they developed for potential inclusion in the CRE websitePostponed Young Career Event “Exposing Research to the Early Career” due to scheduling difficulty
-The purpose of the event is to educate those within their early career on the research field. The presentation will provide an understanding of what ‘research’ really means, and the various facets of research and how it differs across these facets (i.e., agency, media, vendor, sales, etc.). Our aim is to communicate that research is not one-dimensional and can be found in any and allsectors.
-2016 Timing TBDSuggestion to budget for 2016 internships – to help build diverse pipeline of talent for Research Continue working the “Growing the Media Research Profession” with Gary Corbitt
BOARD ELECTIONS
NEW BUSINESS