How TubeMilestone calculates its numbers
Thumbnail & title CTR tier
We don't predict a click-through percentage — nobody can, before a video is published, with any real accuracy. Instead, the AI looks directly at your thumbnail image and title together and rates them low, medium, or high based on characteristics with a documented relationship to CTR: visual contrast, text legibility at small size, presence of a face, emotional signal, and whether the thumbnail and title promise the same thing or contradict each other.
This is a qualitative critique, not a numeric forecast. Treat "high" as "nothing obvious is working against you," not as a guarantee of performance.
Benchmark — "similar-sized channels"
YouTube's API doesn't expose a peer-benchmark endpoint, so we build one at request time: we search YouTube for videos published in the last 60 days that match the topic of your own top-performing video, then look at which channels made them. From those, we keep only channels with a subscriber count between 20% and 500% of yours (minimum 500 subscribers) — close enough in size to be a fair comparison, not the biggest channel in the niche skewing the average.
We average views-per-video across whatever peer channels that search turns up (typically a handful, sometimes more). If the search finds no real peers — a narrow niche, an unusual topic — we say so explicitly and skip the comparison rather than substitute your own numbers or invent one.
Sponsorship rate estimate
This is the one number on this page that isn't a fixed calculation. The AI is given your real engagement numbers (views, subscriber count, niche) and produces a ballpark integration and dedicated-video price range using general knowledge of how sponsorship rates typically scale with those numbers — the same reasoning a human consultant would use, not a lookup table.
Treat it as a starting point for a negotiation, not a quote. Actual rates vary by niche, audience quality, and what the brand is asking for. Always sanity-check against what similar-sized channels in your specific niche are charging before you send a number to a brand.
Comment sentiment
The AI reads up to 24 real comments from your top-performing videos and gives an honest read of what percentage sound positive versus negative or neutral in tone. This is a judgment call on tone, not a keyword-matching script — it's shown free to every user specifically because it's meant to be a genuine read, not something we'd inflate to look better.
What we don't do
- We never show a benchmark number when no real peer channels were found — the report says so explicitly instead of guessing.
- We never invent a "trending topic" without a real engagement pattern in your own videos to point to.
- Revenue/RPM/CPM figures only ever come from a screenshot you upload of your own real Studio data — never estimated from scratch.
See a real report →