Most hashtag advice is a guess dressed as a strategy
FleetClip measures instead. For each candidate phrase it looks at what currently ranks, how many separate channels hold those results, and how recent they are — then decides whether an account your size could realistically get in.
What gets measured
Median views of what ranks
The level a new video has to reach to belong on that page. The middle result, never the average — one runaway hit would otherwise declare every phrase hopeless.
How many channels hold it
A phrase where three channels own every top result is closed, even when the view counts look reachable.
How recent the ranking is
Whether the results are still moving, or settled years ago.
Which wording is winning
Phrasings common to the top-performing titles and absent from the rest — mined from what actually ranked, not from a keyword tool’s suggestions.
Why difficulty scores are the wrong answer
A fixed difficulty score
A phrase is rated 68/100 for everybody. But a query whose top videos average 40,000 views is a wall for an account that gets 500 and an easy win for one that gets 200,000.
Judged against your reach
FleetClip compares each phrase to what your own account actually achieves. The same phrase can be an opening for one account and a wall for another, and it says which.
It stays quiet when it has nothing to say
A topic can be entirely locked up, and that is a real finding rather than a failure. When nothing is genuinely reachable, the research says so instead of offering phrases that will not work.
The same applies to the writing: researched phrases are used where they honestly describe the video, and dropped where they do not. A title assembled out of keywords loses more views to being skipped than it gains from matching a search.
Questions about this
Where does the ranking data come from?
YouTube’s public Data API — real search results and their real view counts. It is used as a proxy for the other platforms, and FleetClip says so rather than implying it measured TikTok.
Does this run on every video?
No. It runs per topic on a weekly cadence, because it draws on a shared daily API quota that every account on the service shares. Running it per video would exhaust that quota within a day and leave every topic without research.
Try it on one topic
Ten videos a month on the free plan, three connected accounts, no card. Enough to see whether what it makes is worth posting.
Start free