Facebook Your Algorithm: Control Your Feed in 2026

For years, Facebook decided what you'd see based on what you clicked, how long you watched, and which posts you tapped. You could like, comment, or scroll past — and the algorithm would slowly figure out what to feed you next. It was a one-way relationship.

Meta just changed the conversation. On October 13, 2026, the company announced "Your Algorithm" on Facebook: a new set of AI-powered controls that lets you tell Facebook, in your own words, what you want to see more or less of in your Feed and Reels. Instead of waiting for the system to guess your mood, you can simply say it.

If you've ever wished you could send a quick note to the algorithm — "fewer outrage-bait posts, more DIY woodworking videos" — this is now, in practical terms, possible.

What Your Algorithm actually is

Your Algorithm is a dedicated preference space inside Facebook where you manage the topics that shape your recommendations. You describe your interests in plain language — a sentence or two, like talking to a person — and Meta's large language models interpret the intent, context, and nuance behind what you wrote. Those written preferences then feed directly into the recommendation system that ranks your Feed and Reels.

This is a real departure from the old setup. Previously, Facebook offered indirect controls: hide this post, snooze that page, pick from a list of broad topics. All of them required you to hunt through menus or manage rigid checklists. Your Algorithm replaces much of that friction with a conversational interface. You don't curate lists; you just explain what you like.

Meta has also been moving in this direction for months. Since December 2025, the company has used text and voice conversations with Meta AI to personalize recommendations across Facebook and Instagram — chat with Meta AI about hiking, and you might start seeing trail Reels and outdoor groups. Your Algorithm turns that personalization mechanism into an explicit, user-facing control.

How to use it

Accessing the feature is deliberately simple:

1. Tap the three-dot menu on any post in your Feed or Reels.
2. Choose the Your Algorithm option to open the preference space, where you can add, remove, or edit topics in your own words.
3. Give feedback on individual posts by tapping "Interested" or "Not interested." Facebook may then ask follow-up questions — for example, what specifically you liked or disliked about that post — so the adjustment lands on the right detail.

The changes don't take effect overnight. Meta says written feedback can reshape your recommendations within the same browsing session, so the post stream should start responding while you're still scrolling.

A few examples of the kind of instructions the system is designed to understand:

• "Show me more cooking tutorials and fewer political debates"
• "I'm planning a trip — give me travel videos for a few weeks"
• "Not interested in viral prank clips"
• "More posts from my local running community"

You can also review and remove the interest topics Facebook has already built up for your account, giving you a rare look at what the algorithm thinks it's learned about you.

Why Meta built it this way

Behind the scenes, this is Meta folding large language models directly into the ranking stack. The models read the freeform text you write — including the subtle context around it — and combine that signal with the traditional engagement signals Facebook has always used: likes, comments, shares, watch time, and thousands of smaller ones.

The philosophy is straightforward: natural-language feedback is higher-fidelity than a like. A like on a cooking video could mean you like the creator, the recipe, the editing, or the background music. Writing "more recipe videos like this" removes the ambiguity. Meta gets cleaner training signal, and you get a feed that converges on what you actually meant.

There's a broader trend here too. Users across every social platform have grown skeptical of opaque feeds. Regulators in the EU have pushed for algorithmic transparency, and Meta's "Dear Algo" experiment on Threads — which let users reshape their feed for three days by posting a message starting with those words — was a trial balloon for giving people explicit influence. Your Algorithm on Facebook makes that influence persistent rather than temporary.

The catch: availability and limits

A few honest caveats before you go hunting for the feature:

• It's rolling out gradually. The October 13 announcement began a rollout, not an instant global launch. Some accounts and regions don't have it yet — notably, reports indicate the feature isn't available in India on Facebook, Instagram, or Threads. If you don't see the option, you're not doing anything wrong; it's just not your turn.
• Written words aren't the whole algorithm. Your feedback joins — not replaces — Facebook's existing engagement signals. If you say "less sports" but spend twenty minutes watching match highlights, the system will get mixed messages. Behavior still talks.
• AI follow-ups are approximate. Language models are good at intent, but they can misread niche or sarcastic phrasing. Expect some trial and error when you describe taste in words for the first time.
• Ads follow a separate logic. Meta says conversations with its AI assistant can already influence ad recommendations — discuss hiking with Meta AI and you may see camping-gear ads — though the company says it excludes sensitive topics like religion, health, politics, and sexual orientation from ad targeting. Your Algorithm preferences govern Feed and Reels content, not necessarily the ad layer.

How to get the most out of it

If you want this feature to actually improve your feed, a little strategy helps:

Be specific. "More tech" is vague; "more honest gadget reviews under ten minutes" gives the models something concrete to rank against. The more precise the instruction, the faster the feed converges.

Use the follow-up questions. The "Interested / Not interested" flow with follow-ups is the fastest feedback loop on the platform. A five-second answer about why you liked a post is worth dozens of passive likes.

Prune periodically. Interests drift. That renovation-project obsession from March doesn't have to define your feed in October. Revisit your topic list every few weeks and remove what no longer fits.

Combine words with actions. Unfollow pages that spam you, mute topics you've outgrown, and use Your Algorithm for the fine-tuning. The explicit controls work best as part of a routine, not as a one-time fix.

Don't expect miracles on day one. Recommender systems need a few sessions of feedback before the written preferences reliably outweigh years of behavioral history.

The bigger picture

Your Algorithm is arguably the most user-friendly algorithmic control Meta has ever shipped — and a sign of where social platforms are heading. The era of the completely opaque feed is ending, not because companies got generous, but because language models made conversational control cheap to offer. If the experiment succeeds on Facebook, expect plain-language feed tuning to become standard everywhere feeds exist: video apps, shopping apps, music apps, news apps.

Whether it makes Facebook better is a separate question, and the honest answer is that it makes Facebook more yours. A feed shaped by your stated intentions will still serve the platform's goals — engagement, retention, ads — but for the first time, you'll have a vocabulary for telling it when it's wrong. That's a small sentence in a menu, and a big shift in who gets to steer.

If the feature has reached your account, open any post's three-dot menu and try one instruction. The worst case is a feed that looks the same as yesterday; the best case is finally seeing the version of Facebook you actually meant to build.