This Post Will Have More Views Than All Of TV* by Joe Marchese @FortuneMagazine May 19, 2016, 9:40 AM EDT E-mail Tweet Facebook Linkedin Share icons The end of April and first few weeks of May are among the busiest times on the media industry’s calendar. It’s traditionally the upfronts season, when TV networks make their top-dollar pitches to advertisers and agencies. Now it’s also the “NewFronts” season, where digital publishers do the same. But this year we in the media world are falling prey to some really wild spin. In digital, it’s become customary for us to invent audience comparisons that just don’t measure up — particularly when it comes to digital video “views.” And when we do this, we make bad decisions. Let me explain. Last month, I’m sure you heard a lot about this video of a watermelon: BuzzFeed originally hosted this video as a livestream on Facebook, and at the peak of its 45-minute run time, 807,000 viewers were watching at the same time — greater than the total viewership for recent episodes of many basic and premium cable shows, like FX’s “Louie” and AMC’s “Halt and Catch Fire.” The press went nuts. Its message: digital is obviously as big as “TV.” (And, yes, BuzzFeed touted this number at its NewFront.) But wait, there’s more! That 807,000 figure represents the number of viewers concurrently watching that single video at a single moment in time, on a platform where a “view” is counted as three seconds. When we say that an episode of “Louie” has about half a million viewers, that’s for any average minute of the show, not at a single “peak” minute. And there was just that one watermelon video — it wasn’t part of an ongoing series of exploding produce. In contrast, “Louie” averaged 336,000 viewers per episode on average across all telecasts on linear TV (all 826 original and repeat telecasts from Season 1 through 5). This is an estimate of “real usage” — taking into account not just how many people watched, but for how long (in the average minute). Taken an important step further: 336,000 average viewership for 826 episodes? That’s a whole lot more real human attention than 807,000 at a given moment in a single video. Comparing these two is less apples-to-oranges and more apples-to-asteroids. We fall for this media spin over and over again. Like when, a couple of months ago, I pointed out a similar mistake in comparing views from a popular YouTube vlogger to the audience of Game 7 of the World Series (the TL;DR version: “Views” do not equal viewers). Or the similar claim this spring that Taylor Swift is bigger than network TV, based on some Nielsen custom research conducted on behalf of Vevo. The week Swift’s “Bad Blood” video was released, her Vevo videos had 18 million views in the U.S. — more than any network TV show that week with the exception of NCIS: Los Angeles. Wow, right? Not quite. What this actually means is that 18 million Americans clicked on a Vevo-hosted Taylor Swift video during that week. Not just “Bad Blood,” but any of the 63 total Taylor Swift videos that were on Vevo when “Bad Blood” premiered. While some of these viewers certainly watched one or more videos to completion, they also could have closed the video after half a second. Meanwhile, I checked with Nielsen to clarify: 19.45 million Americans watched at least a full minute of that single episode of NCIS: Los Angeles that week. What if we talked about TV content in the same hyperbolic terms that we talk about viral videos? Well, it would look like this: FX’s recent “Hotel” season of the American Horror Story anthology series had a multi-platform gross average audience of 10.78 million viewers. When you add up the total minutes of content across the whole season, it results in a tally of 9,046,244,169 gross minutes watched. Divide that all up into two-second views, and you get 271,387,325,078. 271 billion! When I talk about attention using numbers like this, I sound completely insane. But it’s effectively the same metric that’s being utilized for a lot of stuff on the Internet. And just because you can track this metric doesn’t mean you should. Tracked properly, video usage is not just a function of how many watched, but for how long they watched. Nielsen deserves a shout-out for its current work in building a Total Content Rating methodology, which aims to create a multi-platform metric that would enable comparable, and thus meaningful, tracking across screens and over time. But we’re not there yet, and so what we’re currently calling “views” is complete garbage. While it might be tantalizing to utilize this “views” metric to illustrate viral digital hits completely eclipsing traditional TV, the reality is that by doing so, we as an industry are facilitating a false, meaningless narrative. We’re buying into our own bogus spin. It’s not not BuzzFeed’s fault, and it’s not just Vevo’s fault either. Everyone should frame their businesses in the best possible way for investment. But TV has failed to keep up in the larger number messaging department, making it impossible to have meaningful conversations around human attention, thus avoiding the real paths to excelling as advertisers. And I don’t know about you, but I think most of us in the ad business want to do our job well. Premium television content — and, yes, this “television” content can be on digital-first platforms like Hulu or Amazon Prime — is the kind of creative work that draws a passionate, engaged, and critical audience that values the content enough to temporarily attend to a brand message in order to access it. And, since I’m pretty sure this officially makes me the old media guy complaining about all this newfangled stuff, I suppose it’s time for me to get back to telling those meddling kids to get off my lawn. But before I do, let me say this: None of this has to be this complicated! Work backward from the fact that there are 24 hours in a day, and figure out where people are spending the largest share of that time. You’ll be surprised at what starts making sense, and what doesn’t. Joe Marchese (@joemarchese) is the president of advanced advertising for Fox Networks Group, and founder of true[X].