Nielsen, GWI Unveil Fused Data Product Representing 250M Americans

Even as it transitions its core audience measurement service from its legacy panel to a "Big Data-plus" one, Nielsen this morning announced a new product fusing its national TV audience panel with GWI's (Global Web Index) massive digital consumer research panel.

It's hard to understand if the the new service, aptly named "Nielsen GWI Fusion," is a true data fusion hooking respondent-level data between the two services, because the announcement describes it as "overlapping" Nielsen's national TV panel with GWI USA's data, but the companies said the output will enable media buyers to have a "detailed view of today's diverse consumers with more than 250 million people represented."

It would be nice if they also explained the math behind that, because while GWI may or may not actually reach 250 million Americans (about 75% of the U.S. population), Nielsen's national TV panel includes only 80,000 annual respondents.



However the fused sausage gets made, the companies claim it will enable planners, buyers and advertisers to understand how Americans "think, feel and behave" in a way that "goes beyond demographics.

"Users will benefit from a deeper level of analysis to craft impactful advertising and marketing strategies, from deciding on the most impactful combination of media to reach their audience to showcasing the benefits of media efforts."

Beyond the boilerplate hyperbole, the companies said the new product will also enable planners to evaluate explicit variables, such as:

  • Time spent with media.
  • Duplication.
  • Reach and frequency.
  • Enabling a "holistic view of the media landscape" by leveraging the same audiences across data-driven linear.
The companies cited Publicis Media as working closely with them to help develop the new product, and the agency's Executive Vice President-Research Helen Katz stated that it plans to incorporate the fusion into its existing tools and platforms to "enhance our planning, buying and optimization across audiences."

I asked a Nielsen spokesperson to explain the fusion methodology used explicitly, when Nielsen last released a fused data product with another audience-measurement supplier, and for any examples of data insights generated from this new product that I can share with you. Stand by and I will update those later.

4 comments about "Nielsen, GWI Unveil Fused Data Product Representing 250M Americans".
Check to receive email when comments are posted.
  1. Ed Papazian from Media Dynamics Inc, October 31, 2023 at 10:50 a.m.

    Sounds like they are taking a survey that askes respondents about their mindsets, attitudes, product usage, media usage, etc. and statistically mashing the findings on to Nielsen's people meter panel findings by a statistical simulation process. You could do the same thing with the MRI/Simmons data base. The questions remain---how accurate are the findings and how do you determine reach or duplication patterns across platforms if you don't have relaible information on individual programs, website pages, etc.  where the ads appear?

  2. Joshua Chasin from VideoAmp replied, October 31, 2023 at 12:28 p.m.

    Indeed there's been a Nielsen/MRI fusion for ages (I assume it still exists.) This looks like another one of those. I'm guessing that "representing 250MM people" means, GWI is an 18+ survey that is projectable to the 18+ US population, whoich is about 250MM.

  3. Jack Wakshlag from Media Strategy, Research & Analytics, October 31, 2023 at 1:22 p.m.

    Pete Doe at Nielsen knows Fusion better than anyone in the business. Helen Katz is one of the few remaining great research minds at an ad agency. If they stand behind it, it's a solid product. In addition, they can also explain how to do it in a way we can all understand.  Congrats to them. 

  4. John Grono from GAP Research, November 21, 2023 at 7:40 p.m.

    Agreed Jack.   And Josh, you are probably correct.

    One of my concerns of v. large 'panels' is how recent is the data.   If you are focussing on recent trends etc. a mass of that data is pretty much irrelevant.

Next story loading loading..