The subscribe button has been a tombstone since 2020. Your media plan still treats it like a delivery guarantee.
Last month I pulled the analytics on a channel a brand wanted to sponsor. 300,000 subscribers. The last twelve uploads averaged 5,000 views. The brand manager had it tagged as “top tier” in her planning deck because of the sub count. The actual reach was 1.7%.
That gap isn’t unusual. It’s the median.
I’ve spent this year tracking 2,339 YouTube channels — every creator we’ve considered for client campaigns plus a wide control set. The pattern is consistent enough that I now lead with it in every selection conversation: subscriber count tells you almost nothing about the reach you’ll actually buy.
What the algorithm actually does
When a creator uploads, YouTube shows the video to a small initial slice of recently-engaged subscribers. If that slice watches, retains, and clicks through, distribution widens — first to more subscribers, then to Browse, Suggested, Search. If the slice scrolls past, the upload gets buried and the remaining 95% of the subscriber list never sees it.
The subscribe button triggers a candidacy. Not a delivery. A creator with 300K “subscribers” and a flat curve is sitting on a follower badge attached to a tombstone. A creator with 60K subs on a steep upward trajectory is being actively distributed by the platform itself.
Where this hurts the brand
Three places, in order of cost.
One: the projection in the deck always over-delivers reality
If the media plan models impressions from “300K subs × X% reach,” the math looks fine on paper and falls apart on the campaign report. I’ve watched that exact gap become someone’s problem to explain to a CMO three times this quarter.
Two: the cheapest creator on the plan is often the most expensive
Large channels on a decline negotiate harder because they can see their own numbers softening. Their rate looks like a bargain. It’s a discount on a dying asset.
Three: the creators who would actually move the needle are invisible to the filter
The best-performing channels in our 2,339-creator dataset cluster between 50K and 200K subs on strong upward curves. They fail the “100K+ in our vertical” filter every time. They are the entire reason some campaigns beat their projections — and the entire reason most don’t.
What to swap in the brief
I rewrite the selection brief on every deal that lands here. Two changes do most of the work.
First, change the primary filter. Stop sorting by subscriber count. Sort by 30-day average views and 90-day trajectory. A creator pulling 70K views per upload on an upward curve is delivering more than one pulling 30K on a flat 600K-sub channel — and the first one is usually 30–50% cheaper.
The Media-Kit Read
- Subscriber count: 300K
- “Top-tier” tier classification
- Premium rate card
- Looks safe in a planning deck
- Actual reach: 5K, trending down
The Algorithmic Read
- 30-day average views
- 90-day trajectory (up / flat / down)
- Early-minute CTR + retention
- Consistency of upload + delivery
- 60K subs trending up out-delivers 300K flat
Second, demote subscriber count to a tie-breaker. Two creators with similar recent views and similar trajectories? Then sub count matters as a signal of staying power. Before that point, it’s noise.
The steelman for the old filter, briefly: subscriber count used to be a useful proxy because YouTube used to push uploads to subscribers by default. That mechanism died around 2020. The proxy outlived the thing it was approximating. That’s all.
The takeaway
Treating subscriber count as reach is treating a lagging indicator as a leading one. The platform moved on six years ago. Your selection criteria need to move with it.
Pull your last three creator campaigns. Compare the subscriber count in the media kit against the actual views the sponsored video delivered. What percentage did you actually reach?
If it’s under 20%, you didn’t pay for an audience. You paid for a number.