Commentary

Netflix Metrics Help Predict Show Size, Success

by Wayne Friedman , Staff Writer, October 11, 2018

Netflix can tell whether its proposed shows will be small, medium, or large. TV producers will love that. But what about spreading the love to other media business followers?

Netflix's granular expectations of TV shows helps when you have business analysts worried the SVOD service's crazy high production expenditures -- estimated to be $8 billion and growing -- might be an albatross.

Speaking at a Vanity Fair event in Los Angeles on Tuesday, Ted Sarandos, chief content officer, Netflix, did not go into detail regarding specific projections of Netflix shows, but he offered this:

“What we can help them do is size the project. We have enough years of performance data of different shows -- not perfect, but better than most -- to have the ability to say this is going to be a big show, a medium show, or a small show.”

We can also guess that Netflix -- in picking up formerly discarded or revived broadcast TV networks shows -- could count on broadcast network data for shows such as “Arrested Development,” “Fuller House” and others.

Broadcast TV shows for some time -- to a much rougher degree -- predict how new shows will perform, especially when it comes to particular lead-in programming and specific days-of-the-week viewing. To be fair, those metrics increasingly mean less than they did years ago.

Netflix no doubt has deeper measures. It can see exactly when a subscriber is beginning to watch a episode, when that subscriber abandons it, and where and when that subscriber moves to other Netflix content.

Media business analysts are worried about the high-priced cost of signing on big broadcast TV producers, such as Shonda Rhimes ($100 million over three years) or a Ryan Murphy ($300 million over four years). But Sarandos says it just comes down to modern entertainment math.

“It’s not that different from [broadcast] television. A studio or a network would be paying them [what amounts to] a percentage of their advertising and carriage revenue. We are just paying a percentage of our subscription revenue.”

Nice to see a TV service -- digital, linear, on-demand, or otherwise -- that can closely identify expectations, risks -- and hopefully, flops.

So how about showing us some metrics? Forget about viewership of its shows, which Netflix has long refused to share. Just show us your projected TV-show size algorithms. Interested parties want to know.

netflix, streaming, tv

3 comments about "Netflix Metrics Help Predict Show Size, Success".

Check to receive email when comments are posted.

Ed Papazian from Media Dynamics Inc, October 11, 2018 at 10:57 a.m.
Wayne, predicting the rating performance of new shows based on how similar shows have done in the past has always been a very iffy method to say the least. For decades agency media types have played this game with the TV networks, including in their evaluations reviews of the new series' pilots and test results---when available---as well as whatever other info about how the show's concepts tested the storyline and location plans for yet-to-be-made eposides, guest stars selections, etc. they might gleen from the sellers. The resulting predictions have always been subject to wide variations in accuraccy when compared to what Nielsen actually reported when the programs appeared on the air. Why? Because the producers often didn't supply what was promised or because their evaluations of how the stars and settings would be received by audiences were unrelistic,changing public attitudes about their choices of subject matter or the way the show was positioned, etc. ( I cover this in detail in my recent book, "TV Now and Then" )

I am reminded of an analysis made by J.Walter Thompson's media research folks which compared the rating estimates for a new primetime season's progam lineup made by its "experts" with a much simpler analysis. The latter merely assigned all holdover shows the same rating they had achieved the previous season and all new shows the average rating for all new shows. The simple method, which eliminated all subjective aspects, produced a somewhat better answer than the more complex system which tried to evaluate past rating performance for each show plus all of the supposed intangibles. Both, methods, however, were deemed "accurate" only about 60% of the time. And this was in a much less competitive---and rating fragmented--- climate when the three broadcast networks garnered 90% of the primetime audience.

I doubt that Netflix is doing much better but would be interested in seeing an analysis of how its predictions compare with the actual audience results---of course I know that this will never happen as Netflix wants to keep its ratings a secret.
Reply

Jack Wakshlag from Media Strategy, Research & Analytics, October 11, 2018 at 12:41 p.m.

If all I had to do when I was involved in this years ago was to predict “small vs medium vs large” my life would have been simple. What I can say to Netflix here is “ so what”?

brian ring from ring digital llc, October 11, 2018 at 5:06 p.m.

There you go again, Wayne. Concise, readable, dense w insight. The best in the biz, in my book.