How the TikTok Algorithm Decides Your Reach (For You Feed, Explained)

What is “the algorithm,” actually?

When creators talk about “the TikTok algorithm,” they usually mean one specific system: the recommendation engine that decides which videos land on each person’s For You feed (FYF). TikTok describes it openly. It isn’t a single scoreboard that ranks creators; it’s a per-viewer system. TikTok’s current explainer says it uses interaction signals “to create a prediction score for different videos” — an estimate of how likely a person is to interact with each one.

Part of the magic of TikTok is that there's no one For You feed – while different people may come upon some of the same standout videos, each person's feed is unique and tailored to that specific individual.TikTok Newsroom — How TikTok recommends videos #ForYou

That single idea explains most of what feels mysterious about reach. There is no fixed “reach number” attached to your account. Each video is evaluated on its own, for each potential viewer. This is why one post can do very well and the next — same creator, same followers — stalls. It also means the honest answer to “why did this flop” is usually about signals and eligibility, not a secret punishment. If you suspect something more, start with does TikTok actually shadowban.

Which signals does TikTok actually name?

TikTok publicly groups its recommendation signals into a few broad categories. It does not publish exact weights, and neither will we — any specific percentage you see online (“watch time is 40%”) doesn’t come from TikTok. What TikTok does say is that strong indicators of interest, “such as whether a user finishes watching a longer video from beginning to end,” get more weight than weak ones, “such as whether the video's viewer and creator are both in the same country.”

Two takeaways. First, most of what you control lives in the top two rows: make content people actually finish and interact with, and give TikTok clear information (a real caption, relevant sound, accurate hashtags) so it can find the right audience. Second, resist anyone selling a precise formula. TikTok calls maintaining the system “a continuous process” as it works to “adjust models, and reassess the factors and weights.” Note that the Newsroom explainer dates from 2020; TikTok’s current Introduction to the TikTok recommendation system page describes the same signal-based approach.

Why is reach per-video and personalized?

Because the system scores each video for each viewer, your reach is never a flat line. How early viewers respond — do they watch through, rewatch, share? — feeds the prediction for similar viewers. TikTok doesn’t publish the exact mechanics of how a new video’s audience grows, so treat any step-by-step “test batch” formula as a creator theory. What TikTok does say is that follower count and past hits aren’t direct inputs:

While a video is likely to receive more views if posted by an account that has more followers, by virtue of that account having built up a larger follower base, neither follower count nor whether the account has had previous high-performing videos are direct factors in the recommendation system.TikTok Newsroom — How TikTok recommends videos #ForYou

Practically, this means natural variance is the baseline, not the exception. A slow start on one video is not evidence of a penalty. If your whole account drops at once and stays down, that’s a different pattern worth investigating — see why your TikTok views dropped.

What is the eligibility gate?

Here’s the part most “algorithm hack” content skips. Before ranking signals matter, a video has to be eligible for the For You feed. TikTok maintains public For You feed Eligibility Standards:

We make ineligible for the FYF certain content that may not be suitable for a broad audience.TikTok Community Guidelines — For You feed Eligibility Standards

Examples TikTok gives include reused or unoriginal content posted without creative edits (such as clips showing someone else’s watermark or logo), short clips made from GIFs only, shocking and graphic content, mature content, and dangerous activities or challenges. Its 2020 explainer adds that spam and duplicated content aren’t recommended either.

This is the mechanism behind a lot of what creators call shadowbanning. Your video can have a great hook, clean caption and trending sound and still get little reach if it falls under an eligibility standard, because it isn’t in the recommendation pool in the first place. It isn’t silent, though: TikTok says that if content is made ineligible, “this information will appear in the TikTok analytics tool.” The fix isn’t better hashtags; it’s understanding the standards, covered in TikTok’s For You feed eligibility standards. TikTok’s Creator Academy also publishes a plain-language guide to why a video isn’t recommended.

Is it low views or a penalty?

Put the pieces together and most “low views” cases fall into one of two buckets, neither of which is a mysterious punishment:

• Weak interaction signals. The video was eligible, but early viewers didn’t watch through or engage much. This is normal variance — iterate on the hook and retention. TikTok says a video without many views “doesn’t necessarily mean it broke a rule.”
• An eligibility gate. The content falls under an FYF-ineligible standard, so it was kept out of recommendations regardless of quality — and TikTok says that shows in your analytics.

Telling them apart matters because the responses are completely different. Start with each video’s analytics, then watch whether a drop is one-video (variance) or across many posts (worth investigating). We walk through the exact tells in how to tell if your TikTok reach is limited.

You don’t have to guess. Shadowban Checker is an independent, read-only TikTok reach-analytics tool for creators. It connects to your own account through TikTok’s official login and API and estimates whether your reach is being limited — an honest estimate, not an official TikTok status. A free visibility check compares your recent reach with your own baseline and gives a 0–100 risk score, so you can see whether you’re looking at ordinary variance or something worth fixing.