Alternative Truths: The explosion of streaming video has brought with it many myths
Video has a measurement problem. This was true six years ago, when digital platforms and publishers first started to vie for TV dollars en masse through the Digital Content NewFronts, and it’s true now. There is no unified metric in digital that the industry can agree on and use to effectively compare digital performance with TV. And walled gardens have been erected by tech platforms and even major media companies that make apples-to-apples comparisons even harder. In fact, a unified metric may never arrive, and if it does, it might be imperfect at best.
That’s just one of the many lies the video industry deals with on a daily basis. Here are some cold, hard truths:
Views aren’t viewers
If your video show on Facebook has 100 million views over the course of its first season — that’s great, but it’s not bigger than “Game of Thrones.” Too often, digital media companies have tried to show huge reach by conflating their views with TV viewership — even though Facebook measures views only when three seconds of a video have played and most website video players measure once a video has started.
When HBO reports that the seventh season finale of “Game of Thrones” netted 16.5 million viewers, according to Nielsen, it means that an average of 16.5 million people watched every 60-second portion of the entire hour-plus episode. Until tech platforms and digital media companies are willing to adjust to how viewers are defined on TV, comparing the two makes no sense. Of course, some do this to intentionally obfuscate the fact that digital video typically doesn’t get the type of engagement that top TV networks and programs do. (And to be fair to both sides, TV viewers are measured by Nielsen’s representative panel, which is based on its own leap of faith.)
Nielsen isn’t reality
Nielsen is the currency of TV because the entire industry has accepted it as such — not because it’s the most accurate measurement product in the market. Alan Wolk, lead analyst for TVRev, says Nielsen’s argument boils down to the argument that it “can get an accurate snapshot of the viewing habits of 100 million Americans by looking at the viewing habits of 80,000 people desperate enough to agree to have all their TV viewing recorded in return for a small monthly stipend.”
OTT streaming won’t fix linear TV ratings
TV people like to argue that people aren’t watching less TV programming, they’re just watching it on different screens and outside of the first few days of the live linear broadcast. The thinking, then, goes that once networks are able to accurately measure streaming viewing, they can demonstrate how big an audience for a show is — giving a more complete representation of the kind of big viewership TV still gets.
The problem is that this thinking ignores the number of places people can now catch TV shows online. For instance, if a viewer watches a new Fox show on Hulu’s ad-free tier six months after it has aired on the TV network, would advertisers care? Should it be included in the rating at all? Plus, much of people’s OTT watching is of Netflix’s and Amazon’s original TV series, not necessarily traditional TV fare. OTT alone won’t save traditional TV ratings.
Streaming video still has tech issues
The quality problem in the streaming world — where viewers still have to deal with buffering and other playback issues — is not going to fix itself. Expectations are rising just as fast or faster than quality improvement due to high internet capacity, says Aditya Ganjam, chief product officer of Conviva, a streaming video software and measurement company.
One issue is that the quality of the streams receives the most attention, but an equal importance needs to be placed on the quality measurement of the streams. For instance, there are millions of things that can go wrong when streaming video to viewers, so much so that no one human being can sort through all of their issues on their own, Ganjam says. This requires the use of artificial intelligence that is sophisticated enough to diagnose issues in an effective manner. “This is something that only AI can adequately complete,” Ganjam says.
A video ad isn’t necessarily best
The truth is, we just don’t know if video advertising drives sales. Research studies can be done that demonstrate that a particular video ad created enough brand awareness and intent to drive some people into the store — but it’s mostly guesswork with a dash of hope. (Obviously, this is a problem that’s true of all media.)
That doesn’t mean video isn’t effective. But beyond its ability to raise brand awareness, the data just isn’t there yet.
Subscribe to the Digiday Video Briefing: A weekly email with news, quotes and stats around the modernization of video, TV and entertainment.
More in Media
Companies like Priceline and various Amazon vendors are using large language models to update their e-commerce platforms.
Major AI providers and online platforms have a new agreement detecting and preventing AI-generated misinformation.