What audience attention actually costs

A channel with 300,000 subscribers charges more per post than one with 30,000 — that's obvious. What's not obvious: the price difference almost never matches the difference in reach. A large channel's post reach is often 15–20% of its subscribers; a small one's is 40–60%. The large channel's subscriber signed up two years ago and hasn't opened the feed since.

That's why the first thing we do when selecting is calculate the median views across the last thirty posts, rather than take the number from the pinned post. The gap between the median and the best post on large channels can be fivefold. The seller will show the best one.

Across 340 campaigns over three years, the picture is consistent: channels above 200,000 subscribers have a CPM roughly twice as high as channels under 50,000. But conversion from view to target action is 2.5–3x higher on the smaller ones. The resulting cost per action ends up comparable, and sometimes the small channels come out cheaper.

There's a flip side. Small channels run out. To accumulate a million views on channels of 30,000 each, you need 60–80 placements, which means 60–80 approvals, 60–80 invoices, and three weeks of coordination. On large channels, the same volume takes six placements.

The split should mix both types. Large channels build awareness: a person sees the name, remembers it, then searches for it themselves. Small channels drive action here and now. A campaign on large channels only produces a strong reach report and a weak funnel. A campaign on small channels only hits a volume ceiling within a month.

We typically start with a 30/70 split in favor of small channels by budget and adjust after the first two weeks — once it becomes clear which channel type in the specific niche delivers action at a lower cost.

Audience overlap. Ten channels in the same niche give a combined, say, 600,000 views, but the unique viewers among them will be 350,000–400,000. You'll pay twice for every third person. This isn't platform fraud — it's the nature of the niche: people subscribe to everything in their field at once.

Measuring overlap before buying is impossible, but estimating it is doable: if channels share a heavily overlapping list of advertisers over the past three months, their audiences overlap too. We schedule such channels in different weeks of the flight to avoid burning out the same people with frequency.

And one last thing. The cost of attention rises throughout the year: January and August are about thirty percent cheaper than November. Planning an annual budget at November rates is a sure way to end up a quarter short on volume.