Facebook Is Testing a Video-First App: The Placement Shift Most Advertisers Are Ignoring

Meta confirmed on July 24, 2026 that it will test a version of Facebook that opens directly into full-screen video, with the classic Feed demoted to a second tab. The test starts later this year in "video-heavy" international markets, with the US explored in 2027 and an opt-out for users. If you run Meta Ads, this isn't a UX story — it's a placement story. The ad system already sells Feed and video feed placements separately, which means the re-plumbing happened before the redesign. Here's what we're doing about it at Mintec, and what you should do before the test ships.

What Meta actually announced

On July 24, Meta said it will begin testing an immersive, full-screen video experience that opens when users launch Facebook. The classic Feed remains available — as a tab, not the default. Users in the test can opt out. Meta described the initial rollout as "video-heavy" international markets in 2026, and said it will "explore" bringing the experience to the US next year.

The context matters: Meta's short-form video business is now a $50 billion annual run rate. That's the number that explains everything else. Facebook is not testing a video tab; it's testing whether the entire app should become the thing that already makes the money.

The part most coverage missed: placements are already split

Here's what the product coverage mostly skips. In the ad system, Feed placements (1:1 and 4:5) and video feed placements (9:16) are bought separately. That separation exists now, before the test reaches a single user. As AdMake AI's August changelog documented, Meta is already treating the two surfaces as distinct inventory with distinct budgets.

Read that again: the commercial plumbing changed before the UX changed. That's the pattern Meta used with Stories — the ad formats and placements shipped first, the interface followed, and advertisers who treated Stories as a fad paid for it in CPM later. If the video-first test wins, the impression mix shifts hard toward vertical, and the advertisers who waited for the announcement will be catching up from behind.

Why your creative library is the bottleneck

We've been tracking vertical share in Meta accounts for months now — since March, over 90% of Meta's total ad inventory is already vertical format. The video-first test accelerates what that number already told us: square is a legacy format kept alive by a tab that Meta is actively demoting.

In our own accounts, the cost of ignoring this is measurable. When we ran auto-cropped square assets for a DTC beauty client, we measured a 23% ROAS drop versus natively produced 9:16 and 4:5 verticals. Meta's auto-crop center-crops your frame and clips whatever matters — faces, headlines, products. That penalty compounds in a video-first world where square creative gets cropped and deprioritized in a placement that's bought separately.

The 3-layer response we're implementing

This is what we're doing across the accounts we manage, and it maps to three layers: creative, buying, and monitoring.

Layer 1 — Creative: 9:16-first production.

Every new asset for Meta starts as 9:16. The 4:5 Feed cut is generated second, and 1:1 only gets produced when a client explicitly asks for it. This flips the production order we used for years (16:9 → square → vertical) and it costs nothing extra — the 9:16 master is the primary deliverable and the other ratios are crops from it. If your library is square-first, start cutting vertical variants now, not when the test hits your market.

Layer 2 — Buying: treat video feed as its own line of business.

Because Feed and video feed placements are bought separately, you can control the mix. We're reviewing placement breakdowns weekly and setting explicit expectations with clients: video feed is a distinct surface with its own creative requirements, not a free bonus placement that gets whatever's left. Budget that doesn't have vertical creative doesn't get allocated to video feed.

Layer 3 — Monitoring: track video-feed share drift.

You can't wait for a Meta announcement to know the shift reached your accounts. We track the share of spend going to video feed placements weekly, like this:

The point of the thresholds isn't the specific numbers — it's that drift becomes visible before it becomes a surprise. The advertisers who get caught by platform changes are the ones who only look at the account when performance drops.

What we're NOT doing

We're not restructuring accounts around the test yet. Meta runs these international experiments all the time, and most die quietly. What we are doing is treating the direction of travel as certain even though the destination isn't: Meta's ad system now sells vertical video separately, the app is testing opening into video, and short-form video is the $50B growth engine. Each of those facts is independently true. You don't need the test to succeed to justify 9:16-first production.

This connects to a broader shift we've covered before: the Instagram Reels vs Feed placement tradeoff showed that placement cost behavior rarely matches intuition — vertical formats often deliver lower CPMs but different CPA dynamics, and you have to measure per placement, not in aggregate. The unified vertical safe zones Meta shipped in March are now the minimum bar for production, not the ceiling. And the same AI systems that auto-enhance your ads without asking will happily crop or re-format your square creative for a video-first surface — usually not in your favor.

The bottom line

The video-first Facebook test is the loudest signal in years that Meta's default surface is becoming vertical video. The placement system already changed — that's the part you can act on today. Cut 9:16 masters, treat video feed as its own budget line, and watch placement share weekly. If the test dies, you've lost nothing but a few production hours. If it ships, you're a quarter ahead of every advertiser who waited for the announcement.