On-Demand Television Metrics Guidelines

February 4, 2004

Introduction

The rapid pace of technological innovation is reinventing the dynamics of how both content and ad messaging are created and delivered via the medium of television. The transition from analog broadcasting to digital distribution brings with it numerous messaging opportunities as well as inherent technical challenges for TV-oriented advertising.

After months of earnest experimentation and informal dialogue between relevant players in the television industry and the advertising community, it is clear that more formalized metrics of ad message delivery and performance via evolving platforms are necessary for the overall growth of ad-supported television in the coming years. Nowhere is this more evident than in the nascent area of "on-demand" television (defined below), where programmers, multi-channel providers, technology facilitators, and advertisers have struggled to define common expectations for what constitutes ad messaging units, as well as how the performance of those units are captured, reported, and delivered.

The Advanced Television Committee of the American Association of Advertising Agencies (AAAA) hopes to begin formalizing this dialogue with an initial set of desired guidelines that will enable advertisers and their agencies to begin judging the effectiveness of ad messaging in on-demand television environments.

Created with formal input from key industry organizations such as the Cable & Telecommunications Association for Marketing's (CTAM) On-Demand Consortium, the Interactive Television Alliance's (ITA) Interactive Advertising Guidelines (IAG) initiative, the DiMA Group's Innovation in Digital Advertising (IDiA) Consortium, the Advertising Research Foundation (ARF), and a host of television industry sources, these guidelines are designed to represent a formal public starting point for discussion with video program distributors, cable/satellite operators, technology enablers, and research intermediaries toward the goal of defining generally acceptable key on-demand performance metrics.

Definitions

For the purposes of this document, the term "on-demand television environments" is defined to include:

VOD: Real-time, two-way video-on-demand cable services; OR
NVOD: Time-shifted and/or "near" video-on-demand services via:

Stand-alone hard-drive-enabled digital video recording (DVR) or datacast technologies; OR
Integrated hard-drive-enabled DVR or datacast technologies with cable/satellite services; OR

NDVR: Distribution/headend-controlled VOD and NVOD services facilitated by networked digital video recording technology.

As the development of on-demand television continues to evolve, these guidelines may, necessarily, change over time. They are designed to be a starting point to an iterative process. They will also parallel and/or interoperate with similar guidelines to be issued over time for other "advanced television" platforms, such as enhanced programming/commercials, digital ad insertion, virtual channel/walled garden ad units, linear/non-linear advertising addressability, etc.

The guidelines assume that any data collected and subsequently reported will be held to the highest consumer privacy standards, such as those outlined in the Cable Communications Policy Act of 1984 and the individual privacy statements of cable, satellite, and stand-alone technology providers. The expectation is that any collected and reported data will be in anonymous and aggregated form, and not personally identifiable-unless specifically requested by a viewer in an explicit, direct, "opt-in" manner.

These guidelines are minimum expectations of the advertising and agency communities; improvement and refinement of them over time are assumed and encouraged.

Initial Guidelines

1.) Total gross universe counts by DMA: Actual counts of the following-by Designated Marketing Area (DMA)-to facilitate comparative universe bases from most broad to most narrow:

a.) total HHs passed (cable only)
b.) total HH subscribers (across all platforms)
c.) total HH subscribers breakout by analog and digital (cable only)
d.) total VOD-enabled HHs (cable only)
e.) total DVR/NDVR-enabled HHs
f.) total digital set-top box count (e.g., machine address, MAC address, etc.)

2.) Set-top box verification of ad message unit delivery: A total gross count of ad messaging units delivered to set-top box units, where delivery is defined as the first frame and last frame of the ad messaging unit in both video and audio. This may be supplemented with and/or substituted by other generally accepted verification methodologies such as audio/video "watermark" monitoring that relies on the AAAA/ANA Ad-ID code system. Until separate ad messaging unit insertion is feasible, this gross count is expected to cover an entire video stream where an ad messaging unit may reside, inclusive of any content that may be in the stream.

3.) Ad messaging unit viewing duration and "trick-play" behavior: A breakdown of total gross ad messaging units viewed by time duration and/or mode to ascertain whether ads messaging units were viewed in fast-forward or standard mode, and whether rewinding occurred. Also pause function behavior, in both live and recorded modes. Until separate ad messaging unit insertion is feasible, this breakdown is expected to cover an entire video stream where an ad messaging unit may reside, inclusive of any content that may be in the stream.

4.) Time, day and date identification: Supplemental to gross counts for Guidelines 2 and 3, a breakdown of ad messaging unit delivery, duration and behavior by time-stamp (preferably to the second-by-second level), day and date.

5.) Geographic identifier behavior: Supplemental to gross counts for Guidelines 2-4, a breakdown of ad messaging unit delivery, duration, and behavior by, at minimum, aggregate ZIP+4 clusters.

6.) Video link in/out behavior: Where real-time and recorded telescoping/video linking environments exist, a gross count and breakdown of "click behavior," including number of clicks, time of clicks, and the video environments from which those clicks originated (linear ads, content, etc.). Where virtual channel or walled garden "click-to-video" environments exist (e.g., banners, etc.), a gross count and breakdown of click behavior. A breakout of "after-the-click" video behavior (as described in Guidelines 2-5) is also expected.

7.) Data reporting turnaround: All data must be reported in as timely a manner as is technologically practicable-preferably within 24 hours.

8.) Verification/reporting: All data must be either verified (audited) by or reported through a trusted independent third-party research firm.