• by David Goetzl  @dgoetzl, Joe Mandese  @mp_joemandese, October 27, 2009

Following months of speculation and the inadvertent circulation of draft proposals, the media research industry got its first genuine look at what the TV industry's new Coalition for Innovative Media Measurement really has in mind. And apparently, it wants researchers bidding for its proposals to, well, mind the store. "We want to be clear that we are not only looking for better or more robust versions of current media metrics. We need a clear path toward the results-based metrics that are clearly, and rightfully, being demanded by advertisers," reads the official request-for-proposals (RFP)," posted on the coalition's new Web site, which is labeled as still "under construction.-- not for public use, but which can be found here here.

The full RFP, which is printed in its entirety at the bottom of this story in the event that the link is made private after its been published, goes on to say, "We need metrics that reflect return-on-investment (sales or other appropriate business metrics), and do so accurately and on a sufficiently granular level to follow the fragmented media vehicles that are now emerging. 'Opportunity to see' is an important metric, but only the beginning."

According to the RFPs, CIMM is seeking proposals on two studies: one to evaluate how to take advantage of set-top-box (STB) data; and another to examine how to best track viewing across multiple screens. Research companies can bid on either the STB or "Cross-Platform" study - either in full or in part.

In the document, CIMM asks researchers to explain how their work will produce data that "does not replicate commonly used buying and selling metrics."

Looking beyond ratings, CIMM says it is interested in audience information that includes "purchase data, lifestyle data, psychographics and engagement measures." It also cites "sales-based, behavior-shift and/or attitude-shift data."

Research companies are asked to submit their completed proposals - with would-be costs -- to an NBC Universal executive by email.

NBCU's research chief Alan Wurtzel has been acting as CIMM's de facto managing director, while a search for a permanent head continues.

Also in the RFP, CIMM asks researchers if they can provide data on VOD (video-on-demand) and DVR viewing for both national and local markets. CIMM also wants details on how data about time-shifting behavior via DVRs is collected.

A sign of how long CIMM may have been prepping the RFPs: It asks whether researchers can track addressable advertising - referencing a system linked with "Project Canoe." That organization is now known as Canoe Ventures and has put its targeted advertising initiative on hold until at least early next year.

CIMM also asks researchers how they plan on interacting with the Media Rating Council (MRC) -- the media industry's audience ratings watchdog. The inclusion of a reference to the MRC is telling, because CIMM so far has said it does not plan to compete with Nielsen or other media "ratings" providers, but is strictly focused on primary research initiatives. In August, MRC CEO George Ivie issued a statement saying the group would be interesting in liaising with CIMM.

The RFP also asks what plans research bidders might have for integrating with the back office media processing systems such as Donovan Data Systems and MediaBank, which ad agencies utilize to plan and buy media, and to pay media vendors.

CIMM is a gang of 14 that includes multiple owners of TV networks; several ad agencies including GroupM, Interpublic, Starcom MediaVest Group, and Omnicom Media Group; and three big advertisers: AT&T, Procter & Gamble and Unilever. Each has committed to spending $200,000 over the next two years to seed the two studies.

Research companies that may bid for CIMM funding include Rentrak, TNS and TiVo, and maybe even Nielsen, which is concurrently funding new primary research studies that may run parallel to CIMM's via the Nielsen-backed Council for Research Excellence.

Nielsen recently hinted it intends to join CIMM in supporting studies on how to advance media measurement, announcing last week it would contribute an additional $2.5 million to the CRE, bringing the total amount of Nielsen's funding to $10 million.

Below is the text of the RFP document available Monday:

Request for Proposal

1. Background

The Coalition for Innovative Media Measurement (CIMM) is a group of buyers and sellers formed to promote innovation and develop new industry standards for television, mobile and cross platform audience measurement in the United States. To facilitate this ongoing effort CIMM intends to shape, fund, and operate a small number of pilots with measurement companies and publish research findings.

Members of CIMM include, GroupM, Interpublic, Starcom MediaVest Group, Omnicom Media Group, Disney/ABC/ESPN, CBS, NBCU, Fox, Discovery, VIACOM, Time Warner, P&G, Unilever, and AT&T.

2. Prologue

As buyers and sellers of advertising-supported media, we are deeply concerned that, despite the efforts of some research suppliers, media measure is not keeping pace with urgent business needs. The media landscape is changing dramatically, and the television marketplace is changing with it. We are shifting from a TV landscape to a multi platform video landscape with a currency system in place that is stretched to its limits that only measures television. Due to a variety of factors, efforts to upgrade that system and to grow it beyond the one screen are not moving.

We want to be clear that we are not only looking for better or more robust versions of current media metrics. We need a clear path toward the results-based metrics that are clearly, and rightfully, being demanded by advertisers. We need metrics that reflect return-on-investment (sales or other appropriate business metrics), and do so accurately and on a sufficiently granular level to follow the fragmented media vehicles that are now emerging. "Opportunity to see" is an important metric, but only the beginning.

Among our specific concerns two stand out. The promise of set-top-box tuning data, offering very large samples (approaching census) that can deliver the combination of granularity and reliability we need at low cost, is obvious. However research suppliers have not begun to address STB methodological issues to our satisfaction. Moreover, actual STB data from the field has been limited and difficult to use. We are looking for a supplier or suppliers who can show us how STB data works today, and how in the near future it can be part of the results-based metrics the marketplace demands.

A second concern is cross-platform measurement. We need reliable, granular, results-based metrics that can be applied consistently across all the major video platforms, to fuel the burgeoning interest in multi-platform video advertising plans. Such metrics do not now exist. Instead we have a patchwork of measures, each developed for its own video "silo," which makes cross-media planning and buying difficult if not impossible. We need the best minds in the media measurement field to develop ways in which the multiplicity of video media (including, as a start, TV, internet and mobile) can be reported both separately and on a unified basis.

This is the time for leading buyers and sellers to work together with audience measurement providers to seek innovative solutions and understand the tradeoffs of innovative solutions to media metrics. We are agnostic as to who will supply us with this data; the field is open, but we need these forward-looking metrics, adequate to the high standards of trading and post-evaluation, within the next 3-5 years.

3. Proposals Sought

Two studies are initially contemplated, one to investigate the current and future potential of television measurement via set top box data, and the other video cross-platform measurement. Recipients are invited to bid on either or both of these studies, in whole or in part. While numerous questions to be addressed by each study will be presented here it is recognized that recipients may wish to address only those that are within their competencies. You are invited to contact CIMM c/o the name at the end of this document with any questions or comments.

Set Top Box Data Study

Deliverable: three to six months of actual STB data, to be used for evaluation (not sales) purposes; and responses or analyses that address the following questions. Responses may be framed in terms of what is available now or what you plan to provide in the future. If the latter, a timeline should be included.

Cross-Platform Data Study

Deliverable: a pilot study producing real cross-platform data that reflects exposure of specific video sources on television, the Internet and mobile media. Please include in your proposal a description of the video content measured on or across the platforms that you think would be appropriate for a full understanding of cross platform measurement and reflects different types of content. If you require examples of content from CIMM members, specify what you need. Additional deliverables are responses or analyses that address the following questions. Responses may be framed in terms of what is available now or what you plan to provide in the future. If the latter, a timeline should be included.

Questions Relating to Both the STB and Cross Platform Data Studies

Metrics

Do you report "traditional" metrics, those that are currently used in buying and selling? (e.g. reach, frequency, time spent, duplication, etc.)

What new learning does your data provide that does not replicate commonly used buying and selling metrics? This could include (but is not limited to) purchase data, lifestyle data, psychographics, and engagement measures.

How is this data calculated? What are the effects of various way of calculation?

What metrics can you provide that relate specifically to media ROI accountability?

Do you have sales-based, behavior-shift and/or attitude-shift data?

Can you provide VOD, VCR, DVD, broadband and DVR data? On both a national and local market basis?

If you provide DVR data, what methods are used to account for time-shifted viewing? Does your DVR data reflect standalone units or only DVRs built in to a STB?

How are HD networks reported?

What kinds of devices or place based viewing is excluded (bars, hotels, small sets etc.)

Methodology

What methodological issues do you acknowledge with your data?

What level of precision do you anticipate (seconds, minutes etc), commercials

Can you dimensionalize the impact of these issues?

How are you dealing with these issues?

What methodological research have you conducted to validate your data?

Do you include out-of-home TV exposure? Time-shifted viewing?

Do you plan to partner with other suppliers for demographic or other data? Note: a hybrid service is acceptable.

How do you determine what programs/commercials are being played or watched at any moment, for live, delayed and VOD viewing?

Can you ingest precise network logs? How do you identify content that is not coded or for which logs are not provided.

Can you categorize viewers by media behavior, e.g. non-ad-skippers?

Do you employ a single-source sample, or fusion of different samples? If the latter, please describe in detail how this fusion is implemented.

Do you weight your data? If so please describe the process.

What are your plans regarding engagement with the Media Ratings Council?

Other

Please describe your data delivery system. What are your plans for upgrades?

Can a user overlay an advertising schedule on your data?

Can you follow HH-targeted advertising, e.g. Project Canoe?

Do you have software to analyze commercial vs. program data, including pod, pod position, etc., as well as all other data collected?

Can your data interface with established 3rd party processors such as Donovan and MediaBank?

Questions Specific to the STB Data Study

We realize that no STB samples are as yet projectable. What is your plan for getting to national and local projectability?

What is your estimate of subscriber cost increases as the sample expands? What are the skews of your current sample?

Is your current sample drawn from a census or a sample in a specific geography? If sample is used please explain the process.

Can you provide data that is based on multiple markets, to account for different market demographics and STB technologies?

How do you calculate persons' viewing/demographics?

Can you provide second by second data? If so, is this on a production basis or a special-analysis basis?

Describe your edit rules. This should include how you deal with the following:

Box on/set off

Faulting

Dwell time

How do you deal with non-STB sets in a home?

How do you deal with non-STB homes in your sample area?

Do you draw or plan to draw any other data from homes that provide you with STB data? For example HH demographics, shopping card data, modem data, etc.?

Do you capture interactive (non-tuning) data from boxes that support it (e.g. ad click-throughs, EPGs)?

Have you conducted any parallel studies (e.g. telephone coincidentals) to validate your data, particularly data (such as demographics) not drawn directly from your sample?

What is the time lag between telecast and availability of data?

Questions Specific to the Cross Platform Data Study

Describe your edit rules.

What metrics do you provide on (1) aggregate basis, and (2) on an individual medium basis.

What demographics can you provide?

What is the time lag between exposure (on all platforms) and availability of data?

Regarding the internet component, how do you determine what is on the screen (video and non-video)?

Does your internet component include wi-fi enabled laptops, wherever they may be used?

How do you deal with respondents using headphones?

What are compliance rates by medium?

Can you track how consumers move across platforms?

Can you provide insights on differential impact by medium?

Do you measure all types of video, e.g. shows, minisodes, clips, commercials? Note: all of these can contain commercials.

Do you measure video starts and duration?

Additional Information

Proposal Format

The proposal should contain the following components:

Deliverables: specific data and analyses to be delivered.

Timing: calculated from the time the contact is awarded.

Cost: different options may be presented (e.g. months of data).

Company profile: brief history, current syndicated and custom offerings, executives who would be responsible for this project.

Other Terms

Following the date at which the Coalition receives the Supplier's proposal and pricing, that proposal and pricing shall be binding on the Supplier in all respects for a period of 90 days.

In submitting a proposal, the Supplier understands that the Coalition will determine at its sole discretion which proposal, if any, is accepted. The Supplier waives any right to claim damages of any nature whatever.

This RFP defines specific requirements only. It is not intended to be, nor should it be construed as, an offer to contract. The Coalition will consider each proposal, but is under no obligation to act on any proposal. All submitted proposals shall become the property of the Coalition.

Each Supplier agrees that they shall absorb all costs incurred in the preparation, revision and presentation of any proposal.

When applicable, authorized travel and lodging will be billed at cost.