Snapchat and the Extreme End of the Long Tail

With the annual network TV upfronts upon us, I thought it would be a good moment to share some thoughts about my time at NBC and how the broadcast production model contrasts with UGC platforms like Snapchat.

When I worked for NBC Sports, my job was to learn as much as I could about as many Olympians as I could, then make recommendations about which stories we should feature in our coverage. It was important to pick the right ones because we were limited by two scarce resources:

· Broadcast hours — There are only 15 nights of competition during the Olympics, which limited us to 60 potential hours of primetime coverage (or 40 content hours, factoring in commercials). So, even if we devoted every content minute to profiles, we could only feature 800 of the 10,000 athletes (8%) at a Summer Olympics.

· Production dollars — The cost of producing a single three-minute profile of network quality could easily exceed $10,000 (excluding salaries of full-time staff), so we could not afford to profile every athlete even if we had infinite broadcast time.

The result of operating within these constraints was that we would select only a few dozen athletes to profile on NBC. The athletes we selected were those deemed most likely to prove compelling to an American broadcast audience — the famous ones, the likely gold medalists, the ones who had overcome adversity, etc. Because we could not deliver unique content to serve the unique tastes of each of our viewers, we defaulted to content that seemed to offer mass appeal.

This is the inherent challenge of broadcast television, and mass media generally. Making money through video content requires aggregating large audiences, but large audiences don’t like the same things, so producers elevate their videos with big investments in production to attract the fence-sitters who might not otherwise watch. For example, imagine your mother’s tastes as a viewer. If given an option between watching a home movie of you as a kid, a home movie of Beyonce as a kid, and a home movie of a random, non-famous person as a kid, it’s safe to assume she would be thrilled to watch you, be somewhat interested in watching Beyonce, and feel that watching the random person was a waste of time.

Much of what we did at NBC involved using investments in production and storytelling to elevate an athlete from “random, non-famous person” to something closer to Beyonce. We couldn’t make you care as much as a mother cares about her child’s home movies, but we could potentially convert you to a viewer. The result is broadcast television’s bread and butter: content that is slightly compelling to a huge number of people.

What makes Snapchat’s Stories feature so powerful is that it has inverted this model. Because the service is unconstrained by broadcast hours (cloud storage of videos is effectively infinite) and production dollars (users produce their own content with their own phones), it can afford to deliver content about everyone — even those of us who are completely uninteresting to all but a few dozen friends and family.[1] It is the equivalent of NBC airing profiles of all 10,000 Olympians competing in Rio.

This is not exactly the “long tail” strategy foretold by Chris Anderson a decade ago (and refuted by Anita Elberse in her book, Blockbusters: Hit-making, Risk-taking, and the Big Business of Entertainment). Those interpretations of the video industry still focus primarily on commercial content, for example Netflix’s pioneering use of data to match niche titles to viewers with niche interests. Snapchat has done something entirely different. By commercializing amateur content, it has discovered immense value at the extreme end of the long tail where few competitors had even thought to look.

To underscore this dynamic, it’s helpful to visualize the varying degrees of interest for different video content types. In this simplified example, there are two different types of people in the world: normal people and famous people. And there are also three different interest levels: fascinating, interesting and uninteresting. If you’re a normal person, you are fascinating to people who know you (e.g. your mother) but completely uninteresting to the vast majority of the world that doesn’t know you. If you’re a famous person, you are also most fascinating to people who know you, but you are also interesting to some large number of strangers. Of course, even the most famous people are not known to everyone, so eventually they, too, are completely uninteresting to some portion of the population.[2]

As the arrows highlight, our model at NBC was to produce content about athletes viewers recognized (e.g. Michael Phelps) and, when we ran out of those, to take athletes they didn’t recognize and try to elevate them to “interesting” status with expensive investments in production and storytelling. With user-generated Stories, Snapchat needn’t concern itself with a subject’s mass appeal. The few who are fascinated can opt in by following; the rest can ignore.

This paradigm — bespoke user-generated content hyper-targeted to a tiny but fervent audience — is the opposite of the broadcaster’s playbook: instead of a feature that is slightly compelling to a huge number of people, each Story is hugely compelling to a slight number of people. Because “broadcast” time is infinite and production costs zero, Snapchat users can focus exclusively on the most desirable type of content: first-person stories from people they know. And given how small our Snapchat networks still are relative to our actual networks (for those of us older than 18, at least), at scale, this dynamic suggests an outcome where the supply of fascinating content eventually outstrips our available hours for consumption.

Can TV fight back? Maybe, but not at Snapchat’s game.

What makes Snapchat such a disruptive innovation[3] relative to broadcast television is that it combines both demand-side disruption with supply-side disruption. On the demand side, Snapchat unlocked an enormous new customer segment (the world’s youngest smartphone users) by serving them a low-end product (videos of themselves) that broadcasters had neglected to deliver because their core customers (older people) were demanding only high-end productions. On the supply side, the structure of broadcast organizations makes it impossible for them to copy Snapchat’s playbook: scarce broadcast hours and production dollars necessitates a focus on content with mass appeal, which is why you’ll never see an NBC crew show up to your house to film a 10-second disappearing video of you making eggs in the morning.

With no way to match the disruptor, the incumbents may seek to differentiate themselves with progressively higher-quality content. This would be a sensible strategy but also an expensive one, especially in what may be a fight for share of a shrinking pie.

I for one can’t wait for NBC’s coverage of the Rio Olympics this summer — the first time in a decade I’ll be watching from home instead of working. But I must admit I’m also tempted to follow all 10,000 Olympians on Snapchat and experience the Games directly through their Stories.

[1] This is not a new observation, and could be applied to social networks broadly, but Snapchat can be credited as bringing this paradigm to video.

[2] There is a significant assumption here that content about people we know is more interesting than content about famous people. Of course, people we know and famous people come in all varieties, and not every friend is more compelling to watch than every movie star. However, the example of your mother’s tastes, as well as the increasing share of available human attention captured by Facebook, Instagram and Snapchat, would seem to support this generalized hierarchy.

[3] In the classic academic sense, as defined by Clayton Christensen in The Innovator’s Dilemma.