Why ten channels do not give you ten audiences

A media plan adds up reach in a column: twelve channels at fifty thousand each give six hundred thousand. The post-campaign report repeats the same arithmetic. And only when the cost per action comes in higher than projected does someone ask how many unique people were actually in there.

Where overlap comes from

People subscribe to everything in their niche at once. A reader interested in neobanks follows three finance channels, two investment channels, and one news aggregator. Place ads in all six and you've paid six times for one person and once for the other five.

The narrower the niche, the stronger the effect. In mass categories overlap is ten to fifteen percent; in narrow professional ones it reaches half.

How to estimate before buying

Precisely — you can't: platforms don't share subscriber data. But indirect signals work.

First: who advertised on both channels in the past quarter. A fifty-percent overlap in advertisers means a shared audience — other advertisers already found this out the hard way and are voting with their budgets.

Second: mutual mentions and reposts. Channels that reference each other almost always share an audience.

Third: topic specificity. Two personal finance channels overlap less than two channels about mortgages in a specific city.

How to measure after

A promo code unique to each channel helps here. There can't be duplicates — one person enters one code once — so the sum across codes shows the lower bound of uniques. The gap between that number and the total claimed reach is the portion you paid for twice.

The second method is a "how did you hear about us" survey with free-text input. People write the channel name, and in the responses you can see who lists two or three at once.

What to do with this

Don't avoid overlap — it's useful. The second and third contact with the same person work better than the first; that's the whole point of frequency. The question is whether you're paying for it deliberately or by accident.

In practice: we schedule channels with high estimated overlap across different weeks of the flight. That way one person sees the message three times in a month, not three times on a Tuesday, and the third time doesn't trigger irritation.

Where the limit is

An average frequency above four per person almost always means wasted budget. The math is simple: impressions divided by reach. If you took your reach from the sum across channels, the frequency comes out artificially low by exactly the size of the overlap — and you'll be the last to know you overdid it.