Commentary

Big Data + Explainer Video

I'm not sure if Nielsen has ever aired a commercial before, but it just released a nifty "explainer video" vis a vis YouTube to promote the official rollout of its "Big Data + Panel" measurement system as the de facto currency of the new fall TV season.

The video's hook is the longevity and vitality of its sample-based panel, and the video emphasizes it as a constant going all the way back to the 1969 moon landing, although its panel has been around a lot longer than that.

You can watch the video above to get the broad strokes, but the devil is in the details, and the truth is that while the panel continues to be integral to Nielsen's audience-measurement outputs -- you know, ratings -- those estimates are more modeled than ever before (see my recent column interviewing marketing-mix modeling pioneer Ed Dittus and also read the comments section for more on that).

advertisement

advertisement

The bottom line is that making TV/video audience estimates has always been a lot like making sausage, and while most ad pros probably don't want to think about the hows and whys too much, it's important to know there is more factoring and adjustment being stuffed into Nielsen's casing than ever before.

That said, the TV/video landscape has grown so diverse and fragmented that there probably is no other way to get a statistically viable audience estimate, and the Big Data + panel method is now the industry standard per the Media Rating Counsel's recommendations.

Just remember, they have and always will be estimates of the who, what, when, where and why people are watching their shows and the commercials running inside them.

“Combining the trusted, historical measurement of our Nielsen homes with millions of big data inputs gives us the most accurate numbers ever," gushes Nielsen chief  Karthik Rao in a statement released with the new explainer video.

But if you want more confidence than that statement, or the simplistic video, I just so happen to have a pretty in-depth analysis from none other than Howard Shimmel, a one-time Nielsen exec, but a long-time sell-side researcher who currently a member of the team at DatafuelX, which touts itself as an analytics firm specializing in the "convergent TV landscape."

Shimmel has been conducting a series of analyses validating Nielsen's Big Data + panel service, and his latest paper makes a strong case that it is all of the three things that both the supply and demand sides of the ad biz truly cherish: representative, stable and projectable over time.

"The shift from Nielsen Panel to Nielsen Panel + Big Data represents a meaningful improvement in how TV campaigns are targeted, forecasted, and executed. While forecast accuracy gains are most pronounced in lower-rated areas, the benefits extend across audience types and dayparts, with measurable financial impact and improved operational workflows," he concludes, adding:  "Nielsen Panel + Big data also enables persons-based targeting at scale, removes long-standing limitations in panel methodology, and enhances workflow integration — collectively positioning the industry to better manage complexity and drive more precise outcomes."

You can read the paper in its entirety here.

1 comment about "Big Data + Explainer Video".
Check to receive email when comments are posted.
  1. Ed Papazian from Media Dynamics Inc, September 2, 2025 at 1:05 p.m.

    Joe, the main benefit of Nielsen's new "big data" plus panel system is the stability in its ratings that the huge household sample provides.--especially for very low rated program content. However the vtal viewer-per-set factors that will be laid onto the big data set usage findings will be based on  only 25,000 homes where panel member use buttons to indicate that they, specifically, are "watching".Unless this panel is greatly expanded you will have inconsistencies re age, sex estimates--unless a lot of smoothing is done. Whether this goes to accuracy--or predictability---- is open to question. The two are not the same.

    Last but not least, no matter what statistical "modelling"is done the projected "audiences" per commercial--stable or otherwise--- will continue to be vastly inflated--as before. So all we are rally getting is a far more stable system than we now have and the ability to slice and dice the data more finely. Great--but better than ever? I think that the Nielsen ratings of the 1950s-1970s were probably far more accurate for what they were measuring as they were confirmed by many other surveys of the period. Now, the only confirmation we can expect will be when a seller provides real device usage information for comparisons with Nielsen's estimates on indivuidual shows like the Amazon NFL games.

Next story loading loading..