Nielsen Numbers Don’t Add Up for Media Buys

2016 03 23   Peter Feinstein   TV and Ratings    

I remain a skeptic of Nielsen’s much vaunted (by Nielsen) NPX. I don’t buy the validity of mathematically modeling demographic data.

I’m not a conspiracy theorist; all I do is follow the money. NPX sounds like an authoritative, deeply analytical tool that dramatically improves viewership attribution while simultaneously eliminating costly paper diaries. In reality, it’s a profit play.

It’s Nielsen’s market to make, so why not just say that it wants to bolster profits while relying on sound mathematical models to arrive at provable viewer data? Because it isn’t true. Anytime you see statistical wobbles – i.e., going from flat to submariner depths in one short week – you have to know that the math is just wrong. The kind of swings we’re seeing are reminiscent of the days when Arbitron would roll out monthly diaries and we’d see double-digit swings from one month to the next. And they have all the validity of those bogus Arbitron numbers.

I’m not saying that these outlets aren’t experiencing downdrafts in their viewership; I’m simply suggesting the measurement methodology is deeply and irretrievably flawed, and shouldn’t stand up as the numbers we use to buy media.