VNU, meanwhile, appears to be hedging its bets and in the past several weeks has begun pitching clients on a new research product designed to provide similar data as Apollo, but using a controversial method known as fusion. The new VNU service literally fuses two separate databases--ACNielsen's 60,000 HomeScan household panel and Nielsen Media Research's 5,000-plus household national people meter TV ratings system--to create a single pseudo panel that some researchers believe can generate the equivalent of a true single source panel at a fraction of its cost.
In presentations to clients, VNU's Nielsen unit said the new fusion product could provide TV ratings for specific TV shows for people who purchase products in 112 different advertising categories. While the new product does not currently measure other media such as radio the way Apollo would, Nielsen executives showed some of the results and assured clients that the system was a statistically valid way of achieving Apollo-like results.
"This was not a trial balloon. This was a full-blown pitch," said one of the attendees, adding, "This is a real product that they're out there selling."
The new Nielsen/HomeScan fusion is not the first time VNU has pushed the concept of data integration, which has been a goal since the days when Dun & Bradstreet owned the two disparate units. In recent years, Nielsen created a fusion product with WPP's Kantar Media Research's MARS pharmaceutical product usage database, which was considered the first research of its kind to be sold in the U.S. Nielsen has also struck deals to integrate its TV ratings data with Simmons Market Research Bureau's and Scarborough Research's product usage, media and lifestyle databases.
But the timing of the new pitch has struck some observers as either very odd, or extremely telling, coming as the companies are pushing hard to get marketers, agencies and media companies to support an expensive market trial of Apollo next year.
Following a meeting in New York Wednesday with Apollo "sponsors," executives involved in the pitch said the Nielsen/HomeScan fusion discussion did not come up.
"It's a very different product," said one executive involved in the pitch. "It's a low-cost alternative. Some people want a Cadillac. Other people are content with a Ford."
To be sure, there are many research questions concerning the fusion, which essentially ascribes the consumer behavior of people in one sample to "like" people in a separate sample, and treats them as if they are the same people.
The pitch is also the second time VNU appears to be hedging on a joint venture with Arbitron. Several years ago, Nielsen struck an exclusive agreement with Arbitron to develop a new TV and radio ratings system based on the portable people meter, but Nielsen has been loath to commit to a rollout of the project, and did not help Arbitron recruit its next market trial in Houston. Nielsen and Arbitron also have conflicting views on the rollout of a portable people meter sample. Arbitron wants to introduce it to all of the top media markets. Nielsen does not want to introduce it in the top ten markets where it has introduced conventional people meters. Nielsen has said it would make an announcement about its plans to joint venture with Arbitron on the portable people meter by the end of this year.
In a third quarter earnings statement released early this morning, Arbitron described both Apollo and the portable people meter projects as its key "growth initiatives," and noted that its costs related to those projects have risen nearly 12 percent to $52.7 million in the third quarter of 2005 from $47.1 million in the third quarter of 2004.