Grow your TikTok with real accounts.

TikTok’s For You feed ranks videos, not accounts. Its own explainer lists the inputs — user interactions, video information, and device and account settings — and says a strong indicator of interest carries greater weight than a weak one. It publishes no scoring window and no per-signal numbers. A widely reported look at the internals came in the 2021 New York Times disclosure of TikTok’s internal algorithm document — that piece reported a system tuned for retention and time spent. TikTok describes the system in its own words on its own Newsroom “How TikTok recommends content” explainer. Reading what TikTok actually publishes is how creators work out why one video travels and the next one stalls.

Watch time is the part TikTok is most explicit about. Its explainer says a strong indicator of interest, such as whether a user finishes watching a longer video from beginning to end, receives greater weight than a weak indicator, such as whether the video’s viewer and creator are both in the same country. What it does not publish is a test-audience size, a completion percentage, a multiplier or a time window — so any figure you see quoted for those is guesswork. The honest version is simple: a video people finish gives the system a stronger signal than the same number of taps that bounce, and a view counter tick is not the same thing as a completion.

Likes, shares, comments and accounts followed are all listed by TikTok as user interactions its recommendation system reads. What TikTok does not publish is an order between them, so nobody outside the company can tell you that a save outweighs a like or a share outweighs a save — treat every ranked list of engagement types you have read as invention. The part that holds up: a video that holds attention and pulls a mix of responses gives the system more to read than one that collects a single kind of tap.

Audience geography carries less weight than most guides claim. TikTok lists language preference, country setting and device type among its inputs, then says they receive lower weight in the recommendation system relative to other data points, and it names viewer and creator being in the same country as its example of a weak indicator. Region-matched followers still have their uses — brand-deal pitches that need a US or UK audience, comments that read naturally in your language — but treat them as audience fit rather than an algorithm lever.

Follower count is not a direct ranking factor: TikTok states that neither follower count nor whether an account has had previous high-performing videos are direct factors in its recommendation system. What follower count does decide is access. TikTok’s current monetization program, the Creator Rewards Program, asks for at least 10,000 followers and 100,000 video views in the last 30 days, a creator aged 18 or over, and a personal account in good standing in a country where the program runs. TikTok publishes no per-view payout rate, so every cents-per-thousand figure in circulation is an estimate from creators, not a published number.

The shadowban label is doing more work than the evidence supports. TikTok publishes two things adjacent to it: For You feed Eligibility Standards, under which a video can stay up and still be ineligible for the For You feed while people reach it through search or your profile, and guidelines saying accounts caught in deceptive behavior can have their reach restricted, including in search and the For You feed. It publishes no list of hidden penalties, so most of what creators trade as shadowban lore is inference. TikTok’s own Transparency Center reports publish quarterly enforcement numbers, including the volume of fake likes and fake followers prevented and removed. Read them before you buy anything: TikTok’s guidelines prohibit the trade of services that artificially boost engagement, so no provider can put you outside that. What we control is the sourcing — real, active accounts rather than bot networks — and pacing the delivery instead of dumping it.