May has been a cruel month to the TV schedule.
Comcast's (CMCSA 1.36%) NBC just canceled a host of sitcoms: Up All Night, Guys With Kids, 1600 Penn, and Whitney. Even the acclaimed show The New Normal couldn't make it through its freshman season this year.
Disney's (DIS 2.40%) ABC is cutting too. It axed three shows, including Body of Proof. The drama was in its third season and had averaged 8 million viewers.
But that wasn't nearly good enough. Despite the big audiences it and many other canceled programs had earned, network executives couldn't justify holding up one of the precious few broadcast time slots. So they'll have to try again next season with a fresh batch of mostly doomed contenders.
Netflix (NFLX +0.27%) and Amazon.com (AMZN 3.89%) think they have a better way of turning ideas into hit shows.
Netflix has called the current TV model "ripe for replacement," and Amazon's studio arm is busy testing "a new way to greenlight TV shows." However, the two streamers are trying to get to that bold new future in very different ways.
Rise of the machines
Netflix analyzes terabytes of viewer data from its 30 million subscribers. The company follows everything from member watches and rewatches to pauses and rewinds. Netflix algorithms know which directors tend to attract what subset of its audience, and which types of shows tend to do well with other groups. That mountain of information helps it decide what price to bid during negotiations for existing content. But it also guides the company in choosing which original programming to go after.
As an executive told Wired last year, "We know what people watch on Netflix and we're able with a high degree of confidence to understand how big a likely audience is for a given show based on people's viewing habits."
That's how the company could decide to buy two seasons of House of Cards before seeing even a single pilot episode. And Netflix applied the same analysis in choosing to approve the new season of Arrested Development, giving that canceled TV show a new life.
Power to the people
Amazon is taking a different tack. The e-tailer's video service is currently showing 14 original pilots online, and asking viewers to help decide which ones should become full-blown series.
CEO Jeff Bezos thinks this model will work better because it decentralizes the call to green-light shows. "I have my personal picks and so do members of the Amazon Studios team," he said, "but the exciting thing about our approach is that our opinions don't matter."
Amazon's method has the advantage of gathering actual audience reactions. And it should benefit from taking some of the guesswork out of the process. It might even help build buzz for shows, if enough viewers decide to participate. The numbers so far are encouraging. One of Amazon's pilots, Zombieland (based on the movie of the same name), has attracted more than 5,000 customer reviews.
An unfair match
But the biggest advantage over the networks is one that Amazon and Netflix both enjoy: their on-demand platform. That allows them to launch any number of shows without the need to spend millions to assemble a huge audience for episode No. 1. And while ABC and NBC have just a few prime time-slots to trade content around, their Internet-based competitors have no such calendar constraints.
A string of poor choices on a new show can be brutal to the networks' bottom lines. Both Comcast and Disney reported lower ratings for the first three months of 2013, pulling down advertising revenue and profits for the divisions.
Bottom line
Given those weaknesses in the linear TV model, it seems clear that the hit shows of the future won't be chosen by a handful of network executives only to battle over that coveted 8 p.m. Sunday-night time slot. Now the big question is whether it will be a data- or popularity-based model that works better.
And as for the shows that just got hacked out of the network calendar, it isn't curtains for them yet. They can always move online, where they'll have more time to build an audience, and less pressure to succeed right out of the gate.
