• by Dave Morgan , Featured Contributor, February 18, 2010

After my speech at the Borrell Local Online Advertising Conference in New York City last week, an audience member asked whether I thought the recent introduction of set-top-box viewing data into the television media world would "kill Nielsen." My answer was no, I didn't think that Nielsen was going to be put out of business by the availability of census-based television data. But I was certain that the television audience measurement and ad effectiveness marketplace was going to be changed forever.

Here's my thinking:

National ratings as media currency. Like it or not, Nielsen does an extraordinary job projecting high-level national viewing metrics from a sub-20,000 household sample of homes that watch television. The sample and methodology are opaque, but the company does a good enough job that a lot of very smart folks in the media industry  are still comfortable enough to use it as the primary currency for the expenditure of more than $65 billion a year in TV ad spend.

However, the sample cannot accurately project accurate viewership for small cable networks -- a growing sector in the TV market. Nor is it very good outside of the 50 or 60 largest cities in the U.S. -- a good chunk of America, particularly to those of us who grew up in Small Town, USA. Here, initiative such as those led by Kantar Media (formerly TNS) to provide census-based ratings data will certainly have an enormous impact as they gain adoption.

Advertising effectiveness. Nielsen ratings data might be able to project how many folks in the country watched a major network TV show, but the panel is not large enough to connect that viewership to purchase data at a meaningful level of granularity.

Once again, here is an area where set-top-box-based services from Kantar and emerging companies like TGA Global -- which has created a multi-hundred-thousand "single source" marketing measurement panel -- will make big inroads over time. Marketers will wonder why they pay rates based on a "media currency" when they can measure actual sales lift from individual media placements.

Ad targeting. As we read in the trades yesterday, Comcast and Starcom MediaVest have just announced the results of their advertising addressability trials in Baltimore, which leveraged anonymous set-top-box viewing data to improve the relevance of ads delivered to each household.  They found that viewers who saw advertising directed to their household were 32% less likely to tune away from the ad and program than households that received non-addressed ads.

The trial also demonstrated a 65% greater efficiency in reaching target audiences than non-addressed ads. This is very powerful stuff, none of which is possible without census-level viewing data -- an area where companies like Kantar, TRA and also Rentrak are emerging, with solutions that provide alternatives to relying only on panel-based projections.

To be fair, Nielsen, too, is beginning to offer more census-data based solutions in this area, so I'm sure we're going to see the television-audience-ratings business become a very robust marketplace over the next few years. As Rex Conklin, Walmart's senior director of media, said relative to the Comcast/Starcom MediaVest trials, "Walmart remains committed to challenging the marketplace to improve our ability to deliver the right message at the right time and place to our shoppers."

What do you think? Are you ready to challenge panel-data-only approaches, too?

_________