Commentary
Guessing Your New TV Show's Ratings At Episode 50 Or 100
- by Wayne Friedman , Staff Writer, August 5, 2014
Once a TV show gets going big time, TV executives can make fairly predictable assumptions about where a show is going next. Say a broadcast network TV drama starts off at a nice 2.0 rating among 18-49ers, then in its second episode gets to a 1.8; the third, a 1.6; then the fourth, a 1.7. TV program research executives can pretty much guess where the history of that show will be -- in the near term.
In syndication and cable, TV shows can actually be much more predictable, with many development and financial costs also predetermined.
Debmar-Mercury, a TV production/distribution company that works with TV stations and cable network, has worked some new math when it comes to a different sort of TV formula. Offering up 10 “test episodes” of a potential series to a network, it agrees to specific viewing and other metrics. Should the show hit its goals, the network then agrees to buy the next 90, for a total of 100. It’s “the 10/90 rule.” The network gets a nice programming asset -- as well as good money for the TV producer/distributor.
advertisement
advertisement
Why 100 episodes? That is still the measure of a big TV asset, when a series can eventually run Monday through Friday in reruns after its initial airings. It can also be sold more easily internationally than series with fewer episodes.
Debmar-Mercury started this process with a Tyler Perry original sitcom on TBS some years ago, and more recently with “Anger Management” on FX. This week it’ll start up another one on FX --“Partners,” featuring Kelsey Grammer and Martin Lawrence.
The new formula also means a rush to produce all 100 episodes over a shorter course of time -- not the usual four to five year period, but perhaps two-and-a-half years.
Under this or other new business formulas, you can imagine more complicated forecasting tools will be needed to predict where show will be at the end of each of its cycles -- after 10 episodes, after its first season, and then in later runs.
One might predict for any show that there might be a 10% reduction in ratings for a TV show initially after one or more of a couple of “seasons.” But at the end of its run, might that get to 20% to 40% or more? It depends.
What happens to those metrics in a faster moving digital age? Does the rate of media fractionalization and media multitasking contribute to quicker erosion -- which, like gravity, all traditional TV programs eventually cede to?
One thing for sure, for all the fun of new digital media video platforms, a need for proven content will also grow. Business formulas may vary -- and so to will production costs -- but overall content generation is something TV producers can always rely on.
Wayne Friedman is West Coast Editor of MediaPost. You can reach Wayne at wayne@mediapost.com.
Guessing the ratings for new shows has always been a weak area for TV time buyers and media researchers. Many years ago, Marv Antonowski, then at JWT, had an analysis done that showed if you simply assigned all new broadcast network primetime programs the average rating of all new shows last year you got the same degree of accuracy that was attained by all of the judgmental analysis involved in predicting the show by show outcome. Years later, I duplicated this analysis in my newsletter, "Media Matters", and got the same result. There were simply too many unknowns, especially about the casting, writing, and other factors to make really tight predictions.
Big data and machine learning will "eat" the ratings' guessing game, just as software is "eating" every other industry.
Jinni and Amazon Studio are just a couple of the the earliest signals of what's in store.
Longer form content creation will always be part art, part science, but science is now in ascendance.