5 min read

The 7-10 exposure rule: what ten thousand views actually buys

Seven to ten exposures before a listener acts: the attrition arithmetic run against a 10,000-view post, priced against a 360-release catalogue with a median of 2,700 views.

  • Marketing
  • Catalogue
  • Audience

A post clears 10,000 views inside a week. Two enquiries arrive that month, and neither one mentions the post. On its own numbers the post failed, and the old marketing rule says it was never going to succeed alone: a listener takes meaningful action somewhere between the seventh and tenth exposure. Everything before that is groundwork the view counter doesn't score.

This post runs the arithmetic against that 10,000-view upload, because the rule only earns its keep once you can price it.

A view is not an exposure

The player counts a play. An exposure is a person registering that you exist and filing the name somewhere retrievable. The gap between the two is where the 10,000 disappears.

Run it with round working numbers, stated as assumptions rather than findings. Of 10,000 views, perhaps 6,000 belong to real listening: the rest are three-second bounces and repeat plays by the same handful of people. Of those, perhaps a third register the name at all, because most listening happens while cooking, gaming or scrolling with the tag barely heard. Call it 2,000 people who know a name got attached to what they heard.

Now the rule bites. Most of the 2,000 are on exposure one or two with you, and the rule says nothing meaningful happens until seven to ten. The pool deep enough to act is maybe one in ten of them: 200 people, of whom a few act this month. Single digits, out of 10,000 views.

The part no dashboard shows is where those actions credit. Almost never the post itself. The enquiry cites the beat found three weeks ago, the sidebar placement, the loop that ran eleven times while somebody gamed. The post that 'did nothing' was somebody's touch three, and next month's enquiry was scheduled by it.

What the catalogue says about the arithmetic

The discography page carries the raw counts, so I can price my own catalogue against the rule: 360 releases since 2016, just under 39 million YouTube views across them, a median of about 2,700 views per release. Two hundred and fifty-two of the 360 cleared a thousand views. The interesting number is the median. Nothing about 2,700 views converts anybody on its own, by the same arithmetic as the 10,000-view post above.

What the catalogue did was run concurrently. Between 2016 and 2021 it put out 36, 66, 95, 44, 50 and 44 releases a year. A listener who found one type beat in 2018 didn't get a single exposure and leave: the sidebar handed over three more in the same session, the channel page re-served the same artist searches, and every upload collected exposures for every other one. One search worth of interest turned into four touches in an evening.

The top of the view count reads the same way. The Kenny Allstar freestyle with Tion Wayne sits at 309,409 views, 'Parlé' with One Acen at 276,626. Per view, those numbers convert nothing. As exposure inventory, they mark a few hundred thousand people standing at touch one, and the ones who eventually acted did so on a touch the view counter never logged.

Whether an exposure sticks at all is a separate question, and the repetition post covers it properly: an exposure only counts when the name lands on something recognisable. Ten views of beats that each sound like a different producer teach a listener to skip, efficiently.

Counting touches instead of views

The practical version, for anybody running their own catalogue:

Pick the action first. A lease, an email, a follow. 'Meaningful action' that hasn't been named can't be counted, and uncounted actions make every post look like a failure.

Count backwards from the stranger. If you publish three beats a week and a listener found you once, they sit on touch one of seven to ten. The schedule ahead of them is the campaign. Judge the next six weeks of output as a set, not each post against its own enquiries.

Repeat the search terms. The type-beat format runs a cheap sequence for free: one artist search re-served by the sidebar is a second touch you didn't have to earn, and the third follows the same way. Cadence does the rest.

Score the month, not the post. A post that converts nothing but hands 200 strangers their touch three did work the view counter can't see. When the enquiry finally lands it will cite the wrong upload. The rule predicts exactly that, so let it.

Where the rule is honest about itself

The 7-10 rule describes when action tends to show up. It promises nothing about whether ten touches convert anybody, and it cuts both ways: ten exposures to beats with no signature teach a listener nothing worth acting on, so the touches only compound when the thing repeated deserves repeating. The arithmetic also runs on volume. At a median of 2,700 views a release, the catalogue needed 360 of them before the drip of single-digit actions added up to a career. The rule doesn't shrink that work: it explains why the work had the shape it had.

Publish like the count exists, because it does. Name the action, count backwards, hold the cadence, and let the 10,000-view post be somebody's touch three without demanding it be their touch nine too. Next month's enquiry will cite a different post, and it will be right in a way the analytics never were.