Reported on October 6, 2026, Meta began testing a Reels-first Facebook in India and Indonesia — open the app and you land in a full-screen Reels feed, with the traditional News Feed demoted to a tab you can still swipe to, and an opt-out for the users who want the old experience back. It is the Facebook version of a move Instagram made the day before, first outlined in July by Facebook head Tom Alison, who said Meta would test it in video-heavy markets outside the US and bring it stateside next year. The easy read is 'Facebook copies TikTok again.' The more useful read is about distribution, because Facebook is a fundamentally different machine from Instagram and the shift lands differently here. For twenty years, reach on Facebook meant your friends and your Page's followers — the social graph decided who saw you. A Reels-first Facebook hands that decision to a recommendation engine that serves public video to people who do not follow you, ranked on whether the content earns attention rather than on who you already know. That is the real change: Facebook stops being a network you broadcast to and becomes a discovery feed you compete in, the same logic that governs TikTok and a Reels-first Instagram. This guide takes the distribution angle its two neighbors do not — not the supply-chain operating model of the Reels-first strategy guide, and not the production-capacity read of the Instagram-specific one. It is about what a social-graph-to-interest-graph flip means for who you can reach, why it reopens the largest and most-written-off audience on the internet as a distribution surface, what a recommendation-first feed actually rewards from a cold viewer, and the specific ways Facebook Reels are not Instagram Reels. The honest center is that a Reels-first Facebook is less a feature change than a change in what Facebook is for — and the creators who treat it as a real discovery channel rather than a place to dump a cross-post are the ones it will reward.
On October 6, 2026, Meta was reported to be testing a Reels-first Facebook, starting in India and Indonesia. The mechanics are familiar if you saw the Instagram version a day earlier: open the app and you land directly in a full-screen, recommended Reels feed instead of the News Feed. The traditional feed is not gone — it moves to a separate tab a swipe away — and users included in the test can opt out and return to the old feed-first experience. It is an opt-in trial, not a forced redesign, which is the same cautious framing Meta uses whenever it formalizes a shift its usage data has already made.
The plan was not a surprise to anyone watching Meta. Facebook head Tom Alison first outlined it in July 2026, saying the company would begin testing a more immersive, video-forward experience later in the year in video-heavy countries outside the US, and that it expected to bring the experience stateside the following year. India is the obvious first market: it is Facebook's largest by users, and Meta has said young adults there already spend about 60% of their Facebook time watching video. When most of the attention in your biggest market is already on one surface, opening the app to any other surface is leaving engagement on the table. The test is Meta catching the interface up to behavior that already happened.
This guide is one of three on the Reels-first shift, and the angle matters because they do not overlap. The supply-side operating model — why a video-first feed has no ceiling and how long-form becomes the reservoir that keeps it fed — is the Reels-first short-form video strategy. The production-capacity read of the Instagram version — why a feed that rewards video most rewards whoever can make the most of it — is Instagram's Reels-first experience. This one is about distribution on Facebook specifically, because Facebook is a different machine from Instagram and a Reels-first feed changes something about it that it does not change about Instagram.
To see why this lands differently on Facebook, you have to be precise about what Facebook distribution used to mean. For most of its existence, Facebook was a social graph: the system decided who saw your post mostly on the basis of connection — your friends, the people who followed your Page, the groups you were in. Reach was a function of relationships you had already built. That is why Page organic reach became such a grind over the years; you were competing for space in feeds assembled around who people knew, and non-followers were almost impossible to reach without paying.
A recommendation-first video feed runs on the opposite principle. It is an interest graph, not a social graph. The engine serves each viewer the public video it predicts they will watch, drawn from the entire platform, ranked on signals of merit — watch-through, replays, shares, how quickly people swipe away — rather than on whether they follow you. In that world, your follower count is a weak predictor of your reach. A creator with a few hundred followers and a strong clip can out-reach a Page with a million followers and a weak one, because the engine is scoring the content for a stranger, not checking a relationship. This is exactly how TikTok works, and how a Reels-first Instagram increasingly works. The news is that it is now how Facebook works too, for the surface the app opens to.
That is the shift worth internalizing: Facebook is becoming less a network you broadcast to and more a discovery feed you compete in. The broader move of attention onto mobile short-form video that this sits inside is covered in short-form video on mobile is the default now. The point specific to Facebook is that the distribution logic has inverted — from 'who do I already reach' to 'what will a stranger watch' — and that inversion changes who you can reach, what you should make, and whether Facebook is worth your production time at all.
The strategic consequence that gets missed is who the interest graph lets you reach. Facebook's audience is enormous, older on average than Instagram's or TikTok's, and broad in a way that matters for a lot of businesses — it skews toward people with more purchasing power and more established buying habits than the teenage center of gravity on other short-form feeds. For years, creators treated that audience as unreachable organically, because the social graph gated it: without the followers, you could not get in front of them without ad spend. So Facebook became the afterthought platform, the place you mirrored a TikTok to and ignored.
A recommendation-first feed removes the gate. If the engine serves public video to non-followers on merit, then that large, broad, older audience is suddenly reachable with organic content the same way a younger audience is reachable on TikTok — you just have to make video good enough for the engine to surface. That is a genuine reopening of a distribution surface most creators had given up on, and it arrives at a moment when the other short-form feeds are saturating. The creators who adjust first get the arbitrage: a big audience, newly reachable, before everyone else treats Facebook as a serious discovery channel again. The ones who keep dumping a watermarked cross-post there will watch it do nothing and conclude, wrongly, that Facebook still does not work.
Producing for an interest graph is a different craft from producing for a social graph, and the difference is the viewer. On the old Facebook, the people seeing your post mostly knew you; context was free, and a post could lean on an existing relationship. On a Reels-first feed, the viewer is cold — they do not know you, did not choose to see you, and will decide in the first second or two whether to keep watching or flick past. Everything about the content has to earn a stranger's attention from zero.
Practically, that means the disciplines that win on any recommendation feed now apply to Facebook. The opening frame has to promise something immediately, because there is no relationship to buy you patience. The video has to work on mute — a large share of this watching happens sound-off, so the hook and the payoff need to be legible as on-screen text and motion, not carried by audio a scrolling viewer never hears. Burned-in captions are not an accessibility nicety; they are how a muted stranger follows the clip at all. And the whole thing is ranked on completion and shares rather than on the likes your friends used to leave, so pacing and a reason to send it to someone matter more than they did when your audience was people who already liked you. None of this is new to short-form; what is new is that it is now the price of entry on Facebook.
It is tempting to treat Facebook and Instagram Reels as the same product with a different logo, and for distribution that is a mistake. Three differences change how you should treat the Facebook surface specifically.
Since June 2025, Meta has effectively treated all Facebook video as Reels — the separate long-video and feed-video paths collapsed into one Reels upload surface. A Reels-first home is the logical endpoint of that consolidation, not a separate initiative. The practical consequence for you is that there is no 'post a normal video instead' escape hatch on Facebook; whatever you upload is going into the Reels machine, so it should be framed 9:16 vertical and built for that feed, or it gets letterboxed and buried against content that was.
Facebook Reels are public by default, with a private option only on personal accounts. That is not a footnote — it is what makes the recommendation engine possible. Public-by-default plus interest-graph ranking is precisely the combination that lets a Reel reach non-followers at scale. If the instinct is to manage audience tightly, that works against you here; on a Reels-first Facebook the reach is the public distribution, so the move is to produce for strangers and let the engine run.
Two mechanics shape what you ship. Paid promotion (boosting) only works on Reels of 90 seconds or less, so if a clip might ever be boosted, it has to come in under that ceiling. And Facebook screens uploads for copyrighted audio and video through Rights Manager — automatically after you publish, or held for an hour before it goes live if you opt into the in-app copyright check — so commercial audio pulled from outside the platform's licensed library can get the Reel muted, demonetized, or removed. The upload-mechanics version of all this — the create flow, specs, and checks — is in how to make a Reel on Facebook; the distribution point is that Facebook's plumbing has its own rules and a clip tuned only for another platform can trip them.
A few things keep this from being a clean win, and they are worth saying plainly. It is still a test — an opt-in trial in two markets, which Meta could narrow, change, or walk back, and the US timing is a stated plan, not a shipped feature, so building your entire distribution on the assumption of a Reels-first Facebook today is premature. The reopened-audience argument is real but not free: reaching a cold interest graph is harder than posting to followers, and most cross-posts fail there not because Facebook does not work but because the content was never made for a stranger. Recommendation reach is also volatile in a way follower reach is not — a feed that can hand you a non-follower audience overnight can take it back just as fast, so it rewards consistent supply over one lucky clip. And the interest-graph flip does not erase the social graph entirely; Groups, events, and genuine community on Facebook still run on relationships, and a video-first home does not change that those exist. The accurate framing is that Facebook is adding a powerful discovery surface, not replacing everything it was — and the discovery surface is the part that is newly worth producing for.
The behavior a Reels-first Facebook punishes is exactly the one most tools encourage: treating Facebook as the last checkbox in a cross-post, where a video made for TikTok — often still wearing a TikTok watermark — gets mirrored over and left to die. On a social-graph Facebook that barely mattered, because the clip was never going to reach non-followers anyway. On an interest-graph Facebook it is the difference between reaching a large new audience and reaching no one, because the recommendation engine reads a foreign watermark as recycled content and caps its reach. The fix is to produce for Facebook as a real destination, and Kompozy is built to make that the default rather than an extra chore — it is a full generation-and-publishing engine across 18 output formats and eight social platforms plus blog and email, with Facebook a first-class node in the fan-out, not a bolted-on mirror.
Concretely, the same source becomes native Facebook video instead of a downgraded copy. Clipped Shorts cut your long-form into clean vertical 9:16 pieces with no foreign platform logo baked in, Persona Shorts generate talking-head video from a script for the weeks you have nothing to clip, and every output is auto-captioned for the sound-off scroll and framed for a full-screen feed — which is precisely what a cold, muted recommendation viewer needs to follow a clip they did not ask for. The engine produces video built for a stranger, which is the only kind that travels on an interest graph.
Then it publishes that video to Facebook as a genuine destination alongside the other feeds, on a cadence. Autopilot fans each piece across the platforms on durable workers behind a per-post review gate, so a human still signs off before anything ships — and because recommendation reach rewards consistent supply over a single lucky post, the cadence is the part that compounds. A single Persona Brief governs voice and a banned-word list while HyperFrames keep the look pixel-exact, so the volume you need to stay in a recommendation feed's consideration set still reads as one recognizable brand rather than anonymous output. The boundary is worth stating: Kompozy does not decide whether Facebook is right for your audience, write the hook that makes a stranger stop, or guarantee the engine surfaces you — the merit of the clip is still yours. What it removes is the reason Facebook stayed an afterthought: producing native, clean, captioned video for it was extra work, and now it is the same pass that fills every other feed.
A Reels-first Facebook is being read as another TikTok clone, but the thing that actually changes is distribution. Opening the app to a recommendation-driven video feed moves reach off the social graph of friends and followers and onto an interest graph that serves public video to strangers on merit — the same engine that runs TikTok and a Reels-first Instagram, now pointed at Facebook's enormous, broad, older audience. That reopens a distribution surface most creators had written off, provided they produce for it the way a recommendation feed demands: native vertical video, clean of foreign watermarks, built to win a cold muted viewer in the first second, shipped consistently rather than once. Treat Facebook as a real discovery channel and the interest-graph flip is an opportunity; keep dumping cross-posts there and it will keep looking like Facebook does not work — when the truth is it just started working differently.
Not for everyone. As reported on October 6, 2026, Meta is testing a Reels-first experience that opens the app straight into a full-screen video feed, starting in India and Indonesia. The traditional News Feed moves to a separate tab, and users in the test can opt out and return to a feed-first experience. Facebook head Tom Alison first outlined the plan in July and said Meta expects to bring it to the US next year, so treat this as the direction of travel, not a finished global rollout.
Because it moves the decision about who sees your video from the social graph to a recommendation engine. For most of Facebook's history, reach came from your friends and your Page's followers — people who already knew you. A Reels-first feed serves public, recommended video to people who do not follow you, ranked on whether the content earns attention. That flips Facebook from a network you broadcast to your existing audience into a discovery feed where cold viewers are the reach, the same logic that powers TikTok and a Reels-first Instagram.
Yes, and that is the point of a Reels-first feed. Facebook Reels are public by default (the private option exists only on personal accounts), and Meta's recommendation engine surfaces them to non-followers based on performance rather than on your follower count. On a video-first Facebook, the majority of your potential reach is people who have never heard of you — which is a larger opportunity and a harder one, because you earn it on the merit of the content, not on an existing relationship.
Because the usage data is strongest there and the stakes are highest. India is Facebook's largest market by users, and Meta has said young adults already spend about 60% of their Facebook time watching video. Testing a Reels-first home in video-heavy markets is a low-risk way to validate the change where behavior already supports it before touching the US. It is the same cautious, opt-in pattern Meta used for the Instagram version — formalizing where attention has already gone rather than forcing a new behavior.
For many creators, yes — because it reopens an audience most had written off. Facebook's demographic is large, older, and broad, and for years its organic reach for non-followers was weak, so creators cross-posted to it as an afterthought. A recommendation-first feed changes the economics: clean, native vertical video can now reach cold audiences on Facebook the way it does on TikTok. That turns Facebook from a low-priority mirror into a real discovery destination, provided you produce for it natively rather than dumping a watermarked clip from another app.
A Reels-first Facebook opens the app straight into a full-screen video feed and demotes the News Feed to a tab — tested in India and Indonesia as of October 2026, with a US rollout planned for next year. The deeper change is distribution: reach moves from the social graph (your friends and followers) to a recommendation engine that serves public video to non-followers on merit. That turns Facebook from a network you broadcast to into a discovery feed you compete in, reopening its large, broad audience as a real distribution surface for creators who produce native, clean, hook-first vertical video.
Get started → · ← All guides · Compare Kompozy vs other tools