Reed Hastings won't use data to mess with TV plots, and that's why Netflix will win

By Sarah Lacy , written on May 29, 2014

From The News Desk

I really thought about cancelling my trip to the Code conference at the last minute. That's not meant as a dig at the event. Kara Swisher and Walt Mossberg's "D" used to be one of only two conferences I attended every year. This year, however, I'm about to take off for Nashville for Pando's Southland conference, and I want to spent as much time as possible with my children.

I was sitting in my kitchen Monday night completely torn, having just got my two year old settled from a tantrum. I looked at the lineup again. Do I really need to see any of these people talk?

It was Reed Hastings' name that swung me.

There were many interesting speakers at the event. Jimmy Iovine was outrageous and funny and, of course, newsy. Sergey Brin came across as a guy doing exactly what he wanted and who didn't give a shit-- I mean that as a compliment. That's conference stage gold. Ditto Softbank's Masa Son. Travis Kalanick's bizarre attempts at rewriting history never cease to make my jaw drop. Dick Costolo was dressed like a character in Downton Abbey.

But it was Hastings' talk this morning that really appealed. For my money Hastings is one of -- if not the-- best CEOs in tech today. He's navigated from mailing DVDs in envelopes to streaming to creating great award-winning original content during a 12 year run in full glare of being a public company. He's not perfect, but his rebound from his one huge fuck up -- Qwikster-gate -- was astounding.

As he said on stage, Netflix is a comparatively tiny tech company that is sitting right in the intersection of Hollywood and Silicon Valley, competing with HBO and FX on one level, and Amazon and Apple on another.

And that hints at the reason Netflix is winning: Hastings is that very rare CEO who gets tech and also gets creative.

Iovine talked about something similar the night before with Apple's Eddy Cue. It's why Iovine borderline stalked Apple for more than a decade to buy his company: They get creative. They get artists. That's a big part of Steve Jobs' legacy.

But Reed Hastings doesn't get enough credit for the same thing. Re/Code's Peter Kafka asked him in about a dozen different ways whether Netflix's data would impact what happens on Netflix's original shows. Data nerds love to think about the never-before-seen ability to understand exactly when viewers stop watching. Data, they think, will allow showrunners to tighten the "feedback loop," to make sure viewers keep plugged in.

ITERATE! A/B TEST! It's what the consumer Web and the cloud are built on. Given how much data the company collects, it seems almost incomprehensible that Netflix wouldn't go apeshit with the feedback loop.

And, yes, Netflix is one of the masters of the feedback loop. It is one of the best at studying how users interact with its site, continually tweaking it to solve very real problems of browsing, recommendations, search and the like.

But Hastings couldn't have been clearer: The data will never impact creative. That's simply not how television works.

To illustrate the point, Hastings told a story about how he once told David Fincher about the plummeting of viewers during the brutal dog-killing scene in the opening episode of House of Cards. Fincher was horrified at his glimpse at the feedback loop saying, "Don't ever tell me that again."

Content has a big data problem, but that problem is not about crowdsourcing story lines. It's about targeting the people who actually want to sit through that gory dog-killing scene.  After all, if they don't want to watch Kevin Spacey kill a dog, they sure as shit aren't going to stick around when he (SPOILER ALERT) pushes a girl in front of a train.

Put another way: Data shouldn't serve to make content that absolutely everyone will love. It should be used to find the exact people that an excellent -- even offensive or edgy -- piece of content will love.

This is why Netflix is killing it on content and Amazon isn't.

I've long argued journalism allowed data to almost kill it in the early blog wave. Cheap clicks eroded brand advertising, and an obsession with views led to link bait, SEO, and now Facebook-friendly crap. Pando is an outlier in that we don't allow our reporters to see their pageview information. We don't want anyone to know how a story performs ever. Traffic data doesn't make it great journalism, the words on the page make it great journalism.

At the end of the talk, Kara Swisher asked Hastings how he could possibly compete without getting acquired given the deep pockets of his competitors. He shrugged. "Yeah they have deep pockets but that's never helped anyone." Microsoft had deep pockets, and Apple beat them. Silicon Valley wouldn't function if deep pockets were the killer app.

What wins is looking at the world differently than everybody else.

[Photo credit: Re/Code]