The Facebook algorithm just started taking requests. On October 13, 2026, Meta rolled out a feature called Your Algorithm on Facebook. It lets people tell Feed and Reels, in their own words, what they want to see more or less of.

For social media teams, that is a real shift. Ranking used to lean almost fully on quiet signals like likes, comments and watch time. Now users can talk back, and the system says it adjusts right away.

I am Nitin from the Debugged Pro team. Below I break down what Meta actually announced, what is still unclear, and what agencies in India and the USA should change this month.

What Meta Announced for the Facebook Algorithm

Meta shared the update in a newsroom post titled Your Algorithm, Your Words on Facebook. The company calls them "AI-powered tools" that let users shape Feed and Reels. Here is the core of it:

A Your Algorithm page. Users can add and manage the topics they want to see in Feed and Reels.
A new three-dot menu. Meta updated the menu on every post so these controls are easy to reach.
Interested and Not interested buttons. After a tap, Facebook may ask follow-up questions about what the user liked or disliked.
Free-text feedback. Users can type what they want more or less of, and Meta says recommendations update in real time.
LLMs inside ranking. Meta says large language models now help its ranking systems read the interests, context and intent behind natural language.

Facebook Your Algorithm screen showing Interested in and Not interested in topic lists, Facebook algorithm update 2026 — The Your Algorithm screen lists topics a user is interested in, "summarized by AI," plus topics they want less of. Image: Meta, via Meta Newsroom.

Meta also says the system adapts "within the same session." So a change made at lunch can shape what someone sees a few minutes later, not next week.

Where this came from

This is not a brand-new idea for Meta. As TechCrunch reported on the Facebook rollout, Instagram first launched Your Algorithm for Reels in December 2025. It then extended it to Feed and Explore in June 2026.

Threads has a similar feature called "Dear Algo." With Facebook joining, all three big Meta apps now give users some direct say over recommendations.

How the New Facebook Algorithm Controls Work

From the screens Meta shared, the flow is simple. That simplicity is exactly why it matters, because more people will actually use it.

Open the three-dot menu on any post in Feed or Reels.
Pick Your Algorithm to see a list of topics, based on recent activity and summarized by AI.
Add or remove topics under "Interested in" and "Not interested in."
Or tap Interested / Not interested on a single post.
Answer a follow-up such as "Why are you interested?" by tapping a suggestion or typing a reason.

Facebook algorithm follow-up prompt asking Why are you interested with suggested reasons and a free-text box — After a user taps Interested, Facebook can ask why and offer a box to answer in their own words. Image: Meta, via Meta Newsroom.

Look at the example in that screenshot. A coffee post offers reasons like "I love latte art" and "Into pour-over coffee." Those are narrow, specific interests, not broad categories.

That tells us something useful. The system is trying to learn the exact angle a person enjoys, not just the general topic.

Old Signals vs New Signals in the Facebook Algorithm

Meta has long said that Feed is selected, ranked and delivered by an AI system made of many models. Its personalization explainer also lists controls that vary by region and device. What changes now is the type of input the system can use.

Area	Before	Now
Main input	Behavior: likes, comments, watch time and thousands of other signals	Behavior plus plain-language statements from users
Negative feedback	Hide post, snooze, mute	Not interested, with a reason in the user's own words
Speed of change	Gradual, as patterns build up	Meta says changes apply in real time, within a session
Topic view	Mostly hidden from users	Users can see and edit topics summarized by AI

The big point: stated intent now sits next to behavior. A user who watched three cricket clips by accident can simply say "less cricket." Their feed should follow.

What the Facebook Algorithm Update Means for Agencies

Many agencies still plan Facebook content around formats and posting times. Those still matter. But this update pushes topic clarity much higher up the list.

1. Your content needs a clear topic label

If users pick topics, your post has to be easy to place inside one. A Reel that mixes a product demo, a meme and a festive offer is hard to classify. A Reel about "quick vegetarian tiffin ideas" is not.

Ask a simple test question for every post. If a viewer had to name this topic in three words, what would they say? If your team cannot agree, the system probably cannot either.

2. Negative feedback can now be specific

Before, a user could hide a post. Now they can explain why, in their own words. Picture someone typing "no more loan offers" or "too many giveaway posts."

Meta has not said exactly how much weight these statements carry. Still, it is fair to assume that repeated, clear rejections of a content style will cost reach. Clickbait, repeat promos and engagement bait are riskier now.

3. Results can shift faster

Because Meta says recommendations adapt in the same session, a weak post may be filtered out faster too. Weekly reporting can miss these swings. Agencies may need to look at daily reach and Reels plays for key clients.

4. Niche beats broad

The examples in Meta's screens are narrow: ocean kayaking, grill recipes, videography. For brands, this favors a few well-defined content pillars over a scattered calendar. A Pune bakery that posts about "eggless birthday cakes" every week gives the system a clear signal.

India vs USA: Where Agencies Should Focus

The feature is the same, but the way audiences use Facebook differs by market. Here is how I would approach each one.

For agencies serving Indian brands

Test regional languages. Meta has not said which languages the free-text feedback supports at launch. If your audience writes in Hindi, Tamil or Hinglish, check how topics appear on test accounts.
Watch festive-season content. Diwali and wedding-season offers tend to flood feeds. Expect more users to hit Not interested on pushy sale posts.
Lean on local niches. City, cuisine and community topics are easy for users to name and pick.

For agencies serving US clients

Audit Reels strategy. Many US brands repurpose short videos across apps. Make sure each Reel has one clear subject, not a mashup.
Review community content. Groups and local business pages often rely on Facebook. Keep posts tied to the interest that brought members in.
Track holiday-quarter performance closely. This rollout lands a few weeks before Black Friday and holiday campaigns. Set a baseline now, so you can spot changes.

In both markets, Meta says controls can vary by region and device. Do not assume every user sees the feature on day one.

A 7-Step Facebook Algorithm Checklist for This Month

Here is a practical plan you can run for each client. None of it needs special tools.

List three to five content pillars for each brand, written the way a user would name them.
Tag the last 30 posts by pillar and mark any that fit none.
Rewrite hooks and captions so the topic is obvious in the first line or first two seconds.
Cut repetitive promo posts that users are likely to reject in plain words.
Set a reach baseline for Feed and Reels before the feature hits most of your audience.
Use test accounts in India and the USA to see how Your Algorithm describes your topics.
Ask your community directly what they want more of, and match it to your pillars.

Step six is the one most teams skip. Seeing your brand's niche the way Facebook summarizes it can be eye-opening.

What Product and Software Teams Can Learn

This update is not only for marketers. It is also a clear signal about where recommendation systems are heading. Users want to steer feeds with words, not just with clicks.

Think about an e-commerce app or a learning platform. A small box that says "Tell us what you want to see" can feed an LLM layer on top of an existing recommender. That is close to what Meta describes, at a much smaller scale.

If you build apps for clients, a few practical points apply:

Show users their inferred interests, like Meta does with its "summarized by AI" topics. It builds trust.
Let people edit and reset those interests easily.
Log free-text feedback carefully and be clear in your privacy notice about how it is used.
Measure real outcomes, like session length and saves, not just clicks.

Our team helps businesses plan this kind of feature through our software development consulting services. For more on Meta's recent AI moves, read our breakdown of Meta's Muse Wonder frontier AI model.

What We Still Do Not Know

Meta's post is light on detail in a few areas. Good agencies should be honest with clients about these gaps.

Weighting: how much a typed preference counts against long-term behavior.
Languages: which languages the free-text feedback understands well.
Ads: the announcement covers Feed and Reels recommendations, not ads delivery.
Reporting: whether page owners will see any data on topic matches or Not interested feedback.

Until Meta shares more, treat early reach changes as signals to test, not proof. Avoid promising clients a fixed result.

Key Takeaways

The Facebook algorithm now accepts plain-language input through Your Algorithm and Interested / Not interested follow-ups.
Meta says LLMs now help rank Feed and Reels, and changes apply within the same session.
Topic clarity is now a core skill. Posts should be easy for a user to name in a few words.
Specific negative feedback raises the cost of clickbait and repeat promos.
Rollout varies by region and device, so test in India and the USA before changing strategy.

Want more practical guides like this one? Browse the latest on the Debugged Pro blog for AI and tech updates written for agencies.

About the Author

Nitin Kumar is Lead Project Coordinator at Debugged Pro. He works with product, design and marketing teams on web and app projects for clients in India and the USA. He writes about AI updates and what they mean for agencies in day-to-day work.