Your ad is written for everyone. So it reaches nobody.

01What cohort creative actually is

Say you have an image that did well on your page. Normally you would make a new one. Do not. Multiply this one instead. Change 1 line of text, or 1 small part of the picture, so it names a group. "Late thirties single mother." "Video editor using Premiere Pro." "Clinic owner drowning in front desk calls."

Each of those is a cohort. 15 cohorts means 15 versions of the same ad. Same picture, same design, same promise. Only the name of the person changes.

And yes, "late thirties mother", "late thirties single mother" and "early thirties mother" are 3 different cohorts. They will not perform the same. That tiny gap is the whole point.

1 thing means 1 thing. Change 2 and you will never know which one moved the number.

The temptation, once you are in there, is to also fix the design a bit. Do not touch it. If you change 2 things you have learnt nothing, because you cannot tell which change did the work. The only reason this system works is that it holds everything still except the group you named.

02The free research layer most people never open

This is the part that turns a production trick into real research. Instagram shows you how each slide inside a carousel did. Open your post, go to insights, tap into engagement, and you can see slide by slide.

So your carousel is not just a post any more. It is a survey. It tells you which group leaned in and which group scrolled past, and it does that on free reach, before you have put 1 rupee behind anything.

I found this while going through carousels my team had made, and I have used it quietly ever since. When a carousel goes big, the numbers get loud. One of ours reached roughly 8,600 engagements with around 8,000 shares, and the slide by slide view showed exactly where people leaned in.

03Why this beats a normal ad test

A normal creative test costs money and time. You put budget behind 15 guesses, you wait, and then you learn that 11 of them were never going to work. That is an expensive way to find out something the free feed would have told you.

Running the versions for free first flips the order. The market picks the angle. Then you pay to push something that has already been chosen. The slide that won is not a hope any more. It is an ad you already know works.

That is also why this matters for being found at all. The group that recognises itself in your words is the group that shares you, saves you and says your name later. People are now asking a chatbot who to hire, and a chatbot repeats what it finds written about you. Being specific is how you get written about. I have said more about that in will a chatbot say your name.

04Where the cohort ideas come from

Not from a meeting. Not from sitting in a room guessing. They come from conversations that are already happening around you, in the words people actually use.

One of ours came from a client group chat. Somebody mentioned plantar fasciitis in passing. No strategy, no plan, just talking. That became a cohort angle the same day. The habit worth building is noticing that a specific complaint in specific words is a cohort waiting to be written.

This is also where my own setup earns its keep. Calls and voice notes go into a company brain, and a Telegram assistant pulls the lines back out when I need them. So the phrase a client used 3 months ago is still there when I sit down to write. Nobody has to remember it. It is my team and me talking to one place instead of everybody keeping notes of their own.

05What a good cohort line sounds like

Specific enough that the wrong person knows instantly it is not for them. That feeling is a good sign, not a problem.

"If you own a window installation company doing a million a year, we made this scaling roadmap for you." That is 1 product with a thousand possible versions, and every version reads like it was made for 1 company.

Or this one, from an account we run: "The 40s reality check. At 42, recovery stops being about rest. It becomes about knowing exactly which muscle stopped doing its job, and most people never find out." That is not an ad for a clinic. That is an ad for 1 person, at 1 age, with 1 specific frustration.

If nobody can tell it is not for them, it was not written for anybody.

Then the 2 sides start feeding each other. The free slide data picks the group. You move that angle into paid with confidence instead of hope. Paid comes back and tells you which angle brings the cheapest real result, and you hand that back to the free side to go deeper. Neither side guesses. That back and forth is the actual system, and it does not stop.

06The 2 objections I always get

First one: "Will people notice I am running 15 almost identical ads?" Almost never. Each person only ever sees the 1 written for them. The overlap you are imagining lives in your ads manager, not in anybody's feed.

Second one: "Does this feel like a trick?" No. You are not saying anything untrue. You are saying the true thing in the words that person already uses. That is just clear talking, with the effort put in on your side instead of theirs.

That is a 2 hour job and it will teach you more about your market than a month of guessing in the ads manager. It is also the raw material for a lead magnet built out of something real instead of something invented.

07Questions, answered