| by David L. Smith
|
| In my previous articles in this space, I have talked about the work done by the ARF and individual companies to further the adoption of meaningful reach and frequency standards for the industry. This column will provide an update on work throughout the industry. It is important for all to understand that, while great progress has been made in the last twelve months, much of the work being done by stakeholders involved with reach and frequency solutions has yet to see the light of day. I make this point as some major players recently have asked if we couldn't just settle on one vendor as the standard |  | |||
| and move along. In the following article, I will do my best to make the case for continued development and the robust results we will get if we encourage the process of development rather than declare a "winner" at this time. | ||||
Background
ARF meetings in late 2001 and Q1 2002
discussed a marriage of survey research (UCM or user centric measurement)
from companies like Nielsen NetRatings and comScore Media Metrix with that
of third party ad servers (SCM or server centric measurement) to comprise
the ideal Web reach and frequency system. The reason for this on the face
of it is simple: The UCMs measure sites but not advertising schedules.
They give detailed demographics and site cume potential. The SCMs provide
precise data on which computers the campaign was served to and paired
duplication between the sites relative to the ads served, coming up with
an estimate of total reach. However, they do not report out demographics
and cannot accurately parse the information for the U.S. only. It was
postulated by the ARF that a blend of the two databases could yield the
ideal reach and frequency from both a planning or pre-buy standpoint and
for strategic media planning or media mix. In addition, as discussed in my
January article [link to agency/davesmith_4apps.asp], use of SCM data will
enable tracking of R/F for campaigns as they are running and permit
optimization efforts against either reach or frequency depending upon
campaign goals. This will roll-up to a number, which can be used for
post-analysis to compare to the pre-buy or planned R/F.
While we are waiting for the penultimate R/F system, several efforts at UCM-only reach and frequency systems have been released, including WebRF from IMS utilizing NetRatings data and WebPlan from Telmar using comScore data. In addition, a first effort at a combined UCM/SCM R/F system has been released by Atlas/DMT using their own SCM data and comScore's UCM data. WebRF+ from IMS is available to agencies that wish to roll up their own SCM data and combine it with NetRatings information.
On another note, the ARF Pilot Study has begun. This Pilot Study is an effort to compare the data gathered by the SCMs when they serve campaigns to the tracking of the same campaigns in real time by comScore Media Metrix and NetRatings. Initial findings from this study will be discussed at the ARF annual meeting in April with more detailed results released at the June ARF Online Media Council meeting.
But there is much that still needs to be done.
Tasks to Accomplish Next
- Tracking/Post Analysis: None of the vendors has developed a methodology for tracking Reach and Frequency while a campaign is in progress with a demographic overlay. The same is true of post analysis. The result being that we are still a ways away from being able to verify whether the reach and frequency actually delivered is that of the plan. This is crucial as the variables can be huge in this medium, unlike any other.
- Data: The biggest issue is availability of data. The software for UCM measurement exists but not all of the data is populated into the software. The newest comScore release should help in this regard. But the industry still needs to push to all stakeholders to encourage that all three platforms (IMS, Telmar and Atlas) have both comScore and NetRatings data so that the agencies, clients and sites can make real comparisons and decide on the best data source.
- Quality of SCM data: The next biggest issue is the availability of quality SCM data. While Atlas has their data up for use by their system, there remain questions as to whether it is the best possible sample given the potential skew towards Avenue A agency data, where Atlas originated. A DoubleClick release of a cume study or an industry consortium of SCMs releasing data (potentially Atlas/Bluestreak/Mediaplex if DoubleClick does not want to do this) is the best possible solution. The leaders in the industry (read major sites) need to make it worthwhile financially for DoubleClick to do a cume study that is available (for a price) to all in the industry.
Other Factors
- Young-Bean Song from Atlas has made a presentation, reinforced by previous data from DoubleClick and Leslie Wood Research, that site duplication and campaign duplication are different. This is why we need the SCM overlay. Without this overlay and/or without a way to do a post analysis of the planning tools from the media research software companies, there is no way to verify whether the R/F's we are getting are correct or not. Also, as stated above, in addition to planning data, we need to be able to do post analysis against the same demos (post programs from the SCMs right now have no demo overlay) as well as track the accumulation of R/F by demo while it is running, making it possible to tweak the campaign if reach or frequency goals are not being achieved.
- A cautionary note that vendors and users must make sure to compute GRPs and R/F on a total U.S. population basis. Some systems have the ability and even default to Internet population for these computations. The danger here is that the results will not be comparable to offline GRPs for the branding folks. This should be on anybody's list of best practices.
- It is important to note that frequency-capping aspects touted by the R/F and third party ad serving vendors cannot be executed on a campaign-wide basis without site cooperation (a universal cookie would help with this). While we cannot frequency cap with traditional media, it is one more thing that can set the Web apart from other media types.
- We need clear industry definitions of how granular R/F should be: broadcast model or print model. (I come down on the print model side as it gives so much more in the strategic targeting.)
- I have a concern that the data that is being provided to date is like the rating data, all the most recent month. And yet, if I am planning a campaign for summer, I want to see accumulation patterns for last summer, etc.
- The fact that work/home universes are not merged yet for Nielsen is a big issue. We need all universes in one query.
- The ideal final tools will have the ability to pick types of sites from a strategic standpoint and then have an automated R/F optimization program available. Data should be available not only on overall R/F but from a quintile standpoint too.
| David L. Smith, President and Media Director of Mediasmith, Inc. in San Francisco is a nationally known expert in the areas of new media application, media strategy and media planning. A thirty-seven year veteran in the advertising media management arena, Smith has a major involvement in national committee work to establish and refine standards in metrics, business practices and financial issues for Interactive advertising with organizations such as the AAAAs, IAB, OPA and the ARF. He currently chairs the Online Media Council for the ARF. |
