Commentary
TV Needs Better Forecasting, Maybe Some New 'Currencies': Shimmel
- by Wayne Friedman , Staff Writer, April 21, 2022
The TV advertising industry needs better forecasting, says Howard Shimmel, head of strategy for datafuelX, president of Janus Strategy and Insights, and former chief research officer at Turner
Broadcasting.
This will help produce more accurate audience estimates and guarantees-- and more importantly, can address major criticism from the likes …
2 comments about "TV Needs Better Forecasting, Maybe Some New 'Currencies': Shimmel".
Read more comments >
Next story loading 
Wayne Friedman is West Coast Editor of MediaPost. You can reach Wayne at wayne@mediapost.com.
Wayne, I don't get this. Even if we move to a 100 million ACR set reporting base, instead of Nielsen's 42,000 people meter homes TV ratings will still be fragmented, hence less predictable---wont they? Also, sellers usually don't guarantee individual show ratings, they guarantee total schedule delivery ---all shows and episodes in the buy----over certain time frames---often a quarter. Accordingly the ups and downs in their guesstimates tend to cancel eachother out.
It is true, however, that certain sellers deliberately over estimate their future audience tonnage delivery--this is very common for new, nationally syndicated TV shows with no track record in the Nielsens and for very low rated cable channels which can easily create many make good spots by simply adding to their commercial clutter. But if a major seller---- say a CBS TV network, for example-----averaged only 1.5 million adults aged 18-49 during the current season in prime time and it claims that this will double next season I would say that it's the time buyer's duty to challenge this assumption and get involved in a show by show future rating debate with the sellers--before simply accepting such a deal.
One final point. Most TV time buys are still negotiated on those ancient umbrella "demos"--- adults aged 18-49 or 25-54. These once constituted half or sometimes more of the average minute "audience" but now they represent far less---in some cases only 20-25%. Obviously, if the sellers switched to a more sensible audience guarantee base---like adults 18+----which reresents 80-85% of their viewers, their rating estimates would probably be less erratic. Such a move would be no problem for advertisers as 18-49 and 25-54 deals have nothing to do with targeting. Most of the viewers are 50+ like it or not.
Panels aren't the bad guys they're made out to be.
If one company had TV data for every single person in the country, panels would be moot. But that's not the case. Big data (from MSO, SmartTV, or streaming platforms) is usually narrow: Yes, 30M households is a big number, but it's only one in every four US households. Who knows how the rest of them watch TV.
So, two things can happen: big data companies can somehow band together and cover the whole country in all its diversity - in which case a panel will be needed to normalize discrepancies in measurement methods; or big data companies, together, can only cover a portion of the country - in which case a panel will be needed to tell the full story.
Either way, like it or not, panels are here to stay.