Your Audience Can Now Mute Your Whole Topic.

Reach used to be something you earned or did not earn.

It is increasingly something a viewer grants. Facebook, Instagram and TikTok have each shipped a control that lets a person tell the ranking system which topics they want more of and which they want less of, and the direction of travel is one way.

None of these is a kill switch, and none is available to everyone yet. But they change which accounts are safe.

The account that survives a topic dial is the one whose subject a stranger could name in a single word. The account that does not is the one doing a bit of everything, competently, in the middle.

What each platform actually shipped

Three features, three different mechanics.

• TikTok’s Manage Topics sits under Settings, then Content Preferences. It offers sliders across more than ten broad categories, including Creative Arts, Travel, Nature and Sports. It started reaching US users from 30 August 2024, and TikTok broadened its content controls in a newsroom post on 3 June 2025, alongside AI-powered Smart Keyword Filters.
• Instagram’s Your Algorithm launched on 10 December 2025 in the US, on the Reels tab. It shows the topics Instagram believes you are interested in, and lets you boost them, mute them or add your own.
• Facebook’s version arrived with the feed redesign Meta set out in December 2025, which the trade press was still reporting as landing in real feeds through August 2026. Tapping the X on a post now lets a person say which topic is not relevant to them, rather than only snoozing or reporting it.

Nobody has all three of these yet

Rollout state is the most useful fact in this whole story, and it is uneven.

Instagram has moved fastest. By its own dated update notes, Your Algorithm reached Explore globally in English on 15 April 2026, and on 10 June 2026 became one system running across Feed, Reels and Explore. Globally in English is not globally, so a large share of Instagram’s audience still does not have it.

Facebook’s is the least settled of the three. The redesign is still rolling out gradually and mobile first, the in-feed control is described as a test, and Meta has signalled further ranking controls for testing next year.

TikTok has published no country by country availability list for Manage Topics that we could find, so treat it as US first and spreading, not universal.

Frequency, not removal, and why that is the harder version

TikTok is explicit that these settings will not eliminate any topic entirely, but can influence how often it is recommended. Instagram’s framing is much the same: tell it what you want more or less of, and your recommendations adapt.

That is gentler than a block, and for the viewer it should be. For a brand it is harder, because it is invisible.

There is no notification, no metric and no line in your analytics that reads “your topic was turned down this month”. You would only ever meet it as a slow, unexplained decline in recommended reach on an account that did not change what it was doing.

That is precisely the pattern most teams misdiagnose as an algorithm change and answer with more posting.

The follow is the protected relationship

The single most useful sentence in any of this comes from TikTok, which states that the topic settings do not affect the content you see from accounts you follow.

Read that again as a strategist. A person can turn your entire category down and still see you, provided they followed you first. On TikTok, recommendation reach is now something a viewer can dampen. The follow is not.

Instagram has not said whether suppressing a topic can quieten a post from an account you already follow, now that Your Algorithm covers Feed as well as Reels and Explore. That is genuinely unknown, and worth watching.

Either way, the follower stops being a vanity number and starts being the part of your distribution a viewer has already opted into. That is the same logic behind building a narrow community rather than chasing views.

Why the safe middle loses first

These controls operate on categories, not on accounts. A viewer does not turn you down. They turn down the bucket the classifier put you in.

So the only question that matters is which bucket you are in, and whether you sit near the middle of it or clearly at one edge.

An account posting a mix of office culture, product shots, industry commentary and light humour has no edge to sit on. It gets classified generically, and it then rides whatever happens to that generic category across a million people’s settings.

An account that is unmistakably about one thing gets classified as that thing. Some people will turn it down hard. The people who turn it up get a far stronger dose of you. On an opt-in feed, that trade is a good one.

A narrow topic is not a small audience

The objection to narrowing is always the same. We will cut our reach.

Narrow describes the subject, not the size of the room. A channel about restaurant kitchen operations is narrow and has a global audience. A channel about business is broad and has almost no audience, because nobody has ever wanted business in general.

The dial makes this concrete. A person can only turn a topic up if the topic has a name. Being hard to categorise is not neutral any more, it means you are never the thing anyone chooses more of.

The one word test

Take nine of your recent posts, show them to someone who does not follow you, and ask them to describe the account in one word. Not a sentence, one word.

If you get nine different words, or “marketing”, or “a company”, then the ranking system is guessing too.

Run it internally first. Ask three people on your team to write their word down separately, without discussing it. Founders are usually surprised by the spread.

The word you want is something a person could plausibly want more or less of. Sourdough. Padel. Payroll. Rally driving. “Lifestyle” is not a topic, it is a shelf.

Fewer lanes, not more posts

The reflex when reach dips is volume. It is the wrong lever here, because the thing being turned down is a category, and posting more of an unclear category only gives the classifier more evidence of the same unclear thing.

Cut to two or three content lanes, named in the words a viewer would use rather than the words your team uses internally, and hold them for a full quarter.

Three lanes with a recognisable format each will read more clearly than seven lanes at the same total volume. Same effort, sharper signal.

If your team cannot agree on which word it is, that disagreement is the actual problem, and it is what a social audit and strategy engagement exists to settle.

What good looks like

• A stranger can name your topic from the grid alone, without reading a caption.
• Two or three lanes, each with its own recognisable format, rather than seven overlapping ideas.
• Follower growth is a stated objective and gets reported on, not waved away as a vanity metric.
• One person owns the topic definition and is allowed to refuse posts that fall outside it.
• The publishing plan is written in the viewer’s language, not the org chart’s language.

The honest limit, which is that most people never open a settings menu

Overstating this would be a mistake, so here is the counter-argument.

None of the three platforms has published how many people actually use these controls, and we would not assume the number is high. A setting buried three taps into a menu has always been minority behaviour.

The precedent is instructive too. Facebook has offered Show More and Show Less on individual posts since October 2022, and no category of page fell off a cliff because of it.

Facebook’s new in-feed version is the one to watch, precisely because it lives where attention already is instead of in a settings screen. One tap is a much lower bar than four.

Treat this as a slope rather than a cliff. Clarity gets rewarded a little more every quarter, and vagueness gets punished a little more. Neither happens overnight.

What to change this quarter

1. Write your topic in one word, then test that word on three people who do not work with you.
2. Count your content lanes. More than four and you have a positioning problem, not a volume problem.
3. Look at the split between recommended reach and follower reach on each platform, and note which one is carrying the account.
4. Decide what you will stop posting. The lane you keep for internal political reasons is usually the one blurring your classification.
5. Repeat the one word test in ninety days against new posts, and see whether the answer has sharpened.

How NBK thinks about topic definition

Topic definition is an operations problem before it is a creative one. Teams rarely blur their topic on purpose. They blur it because four stakeholders each get a post a week, nobody owns the word no, and the calendar fills with requests instead of a plan.

So we treat topic as part of the operating system rather than a brand workshop: one owner, a written definition, a lane list the calendar has to obey, and a monthly check that what shipped matches what was agreed. That is the same discipline behind running social as a system.

Our team has run always-on publishing at scale, 600M+ views a month, and the pattern holds at every size. The accounts that compound are the ones a stranger can categorise in a second.

Next step

Opt-in reach does not punish small accounts. It punishes unclear ones.

If your output feels busy but your topic is hard to name in a word, start with an audit. The constraint is almost always the definition, not the volume.