In 2016, every digital publisher was told the same thing by Facebook:
video was the future of the platform. Engagement on video posts was dramatically higher than on text posts. Video was where the audience was going. Publishers who didn’t pivot would be left behind.
The publishers that had built editorial staffs, developed distinctive voices, and created the kind of reading experience that an audience came back for specifically made a decision: they would become video companies. They would let writers go and hire producers. They would redirect editorial budgets toward video infrastructure. They would be where the audience was.
The numbers were wrong.
What Facebook actually did.
Facebook’s video engagement metrics inflated average watch time by calculating it only among viewers who watched more than three seconds — excluding anyone who scrolled past — while presenting the numbers as if they reflected the full audience. This made the average watch time appear to be somewhere between 60 and 80 percent higher than it actually was.
In 2019, Facebook reached a settlement with advertisers who had sued over the misrepresented metrics. The company acknowledged the calculation error. The settlement was for a reported $40 million.
The publishers who had restructured their entire editorial operations based on those numbers were not parties to the settlement. They had already fired the writers.
What the pivot actually produced.
BuzzFeed had built its editorial identity on a specific kind of culturally literate list content and, later, on a genuine reporting operation. The pivot-to-video phase produced video that was technically competent and largely indistinguishable from every other publisher’s video. The editorial voice that had made BuzzFeed recognizable was a text voice. It did not survive the format transition.
Mic, which had built an audience on social-first political and cultural coverage, pivoted to video in 2017. By late 2018 it had laid off most of its staff. The audience it had built on text content did not follow it to video.
Vice had built an editorial reputation on immersive first-person journalism and documentary work. The pivot toward Facebook video formats — shorter, social, optimized for the newsfeed — produced content that looked like Vice and wasn’t. The editorial sensibility that made Vice’s best work distinctive required a format and a duration that Facebook video didn’t support.
The pattern across dozens of mid-sized publishers was identical: editorial identity built in text, pivot to a format that didn’t support that identity, audience loss, staff reduction, eventual acquisition or closure at a fraction of previous valuations.
The structural mistake.
The pivot-to-video story is not primarily a story about bad data, though the bad data made it worse. It is a story about editorial decisions being made by distribution logic.
A publisher’s editorial decisions should be made based on what serves its audience, what its voice is suited to, and what it can do better than anyone else. These are editorial questions. They require editorial judgment.
The publishers that pivoted to video answered these questions with platform data: Facebook says video is performing better than text, therefore we should produce video. This is distribution logic masquerading as editorial judgment. The platform’s interests and the publisher’s interests are not the same, and using the platform’s metrics as the criterion for editorial decisions is always a mistake — eventually, sometimes catastrophically.
This is the pattern that repeats through the history of online media. The platform changes the metric. Publishers optimize for the metric. The metric changes or turns out to be wrong. The editorial staff is gone. The audience has moved on.
What the survivors did.
The publishers that survived the pivot-to-video era without losing their editorial identity shared a structural characteristic: someone in the organization had the authority and the standing to say “this is not what we are” and be heard.
Not a stubbornness about new formats. Video is a legitimate editorial format. But the decision to move into video should have been an editorial decision — what can we do in this format that serves our audience and reflects our voice? — not a distribution decision — the platform says this format is performing better.
When editorial authority is subordinate to distribution data, the organization will always optimize for the platform rather than for the reader. And platforms change.
What content strategists should take from this.
The immediate lesson is: Don’t make editorial decisions based on unexamined platform data. Question the metrics before you restructure around them. Understand how they’re calculated. Ask whose interests they serve.
The deeper lesson is structural: editorial judgment needs protected authority. Not veto power over every marketing decision — but genuine authority over editorial decisions, including the decision about which formats and channels are appropriate for the organization’s voice and audience.
A content operation where editorial decisions are made by the algorithm is a content operation that will be restructured every time the algorithm changes.
Which is to say: constantly and without notice, and often without reason.
Jacob Clifton is principal at Clifton Creative.