Meta Tests Video-First Openings for Facebook and Instagram
Meta is testing opening Facebook and Instagram straight into video.
In India, Facebook is testing a launch that opens the app directly into an immersive, full-screen video feed. Instagram, in parallel, is testing an option to make Reels the first screen on open The Verge.
The Facebook test is limited by place for now. Meta said the full-screen experience would go to "video-heavy countries" first, with a possible US expansion in 2027.
The language here is worth noting, because it points to a rollout ordered by how much video people already watch, not by platform age or market revenue.
The Instagram test is an opt-in default with a way back. Meta spokesperson Devi Narasimhan confirmed to The Verge that the Reels-first test is live in a handful of unspecified countries. Users in the test can switch back to the original layout. In the test, Reels turn sound on once a user swipes or clicks into a video.
That audio detail is worth flagging, since it changes how sessions are counted, how much sound-on ad space exists, and what counts as watching to the end.
The tests coincide with more video watching on Facebook. Time spent watching video rose 9 percent globally and 10 percent in the US and Canada, as discussed in Meta's recent earnings call The Verge.
The broader pattern is straightforward, defaults tend to follow attention, and attention is shifting to video.
Meta has tried versions of this start for years. In 2024, it combined Reels, longer videos and Live in one fullscreen Facebook experience, then added a full-screen Video tab that gave Reels a more prominent home. Instagram tested a TikTok-like full-screen vertical feed in May 2022, nearly full-screen videos in June 2022, and ultra-tall 9:16 photos to fit its Reels feed in August 2022. Facebook described a reimagined full-screen video experience for a subset of users in July 2026. Reels was announced for Instagram in August 2020 and launched on Facebook in more than 150 countries in February 2022.
The broader context here is entry-point control. Social apps have two ranking problems. One is what to show. The other is where the session starts. Change the start and the rest shifts, from cold-start signals, or early guesses about intent, to how viewing time spreads, swipe-through rate, and which recommendation system is called first. A Reels-first open favors the interest graph, videos picked from what you watch, and short-loop retention over the follow graph, posts from accounts you follow, and recency. For engineers, that means fewer format compromises on first paint and clearer data on intent.
In my view, the reversible design is the telling part. Letting users switch back lowers the cost of testing and limits backlash, while still giving a clean A/B comparison on retention, session length, and creation. I have watched my own children move past still launch screens straight to video without a pause, and that habit now reads like product specification. The default is catching up to behavior already clear in the numbers.
Looking at what this means for builders, a video-first default centers ad space on vertical, sound-on, full-screen units. That simplifies creative rules and measurement, especially for Reels ads up to 30 seconds long tested in vertical full-screen form, and for 9:16 stills made to hold up in a video feed. It also raises the bar for player speed, prefetch discipline, or loading next clips ahead, and inference latency, or how fast the model picks on launch, because stutter at entry loses viewers. Creators built for feed placement and thumbnail clicks will need a hook in the first second and pull through the loop.
Looking ahead, the open question is not whether short video has earned the default. The numbers already lean that way. It is whether users still want distinct apps with distinct starts, or one continuous recommendation feed under different names. Meta appears willing to learn market by market, starting where video is heaviest, a path that could make discovery faster and creation simpler.
