How to tell what really took off in your niche

This is the method I use to see what really took off: five steps to measure each post against its own account's normal, without likes fooling you. It's what didn't fit in the carousel or the reel. You can do it today, by hand, with your account and a calculator.

Why likes alone fool you

If you sort your niche by likes, the big account always wins. You can't tell what's an ordinary day for it from what really took off, and you end up copying what was already big.

What works is comparing each post to its own account's normal. That's the index: the post's likes divided by that account's normal. A 10× is a post that did ten times what its account does on an ordinary day. The one that took off, took off on its own turf.

And it applies to your account: if you post more and reach less, the total tells you nothing; each reel against your own median does.

The method, in five steps

1. The list. The accounts you want to look at. Start with yours; then those in your niche, few and well chosen.
2. The normal. Write down the likes of each account's last posts, about 30, in a row, without skipping the weak ones. The normal is the median, not the average: sort the likes from lowest to highest and take the one in the middle. A single post that blew up inflates the average; it doesn't move the median.
3. The index. Post likes ÷ its account's normal. A 1× is an ordinary day.
4. The cutoff. Don't look at everything. Sort from highest to lowest index and keep the ones that at least doubled the normal. The rest is routine, not a pattern.
5. The pattern. From the one that took off, you copy the pattern that took off. Never the text.

A full example, with sample numbers

Take a small account, with sample numbers, made up so you can see the math. Of its last 30 posts, the normal comes out to 80 likes. Five of those posts, sorted by index:

• 800 likes · 10× · took off
• 240 likes · 3× · took off
• 160 likes · 2× · took off, right at the cutoff
• 80 likes · 1× · an ordinary day
• 40 likes · half the normal · fell short

Notice what the list doesn't say: how many followers the account has. With a normal of 80, 240 took off; on a big account, 240 can be a weak day. That's why the index compares a small account with a big one, and you with yourself last month.

If none of yours reaches double, don't lower the cutoff to make winners appear. That's data too: measure an account in your niche and see what takes off there.

What the index lets you see

Since it compares each account with itself, it shows what likes hide:

• News is worth less after a week. The same account, the same topic, a few days apart: the second post falls into routine. It isn't won by whoever reports it first, but by whoever tests it.
• Your best month can hide your worst reel. The total goes up; each post's index tells you which ones contributed and which were routine. It's the yardstick that doesn't move when you post more.
• The pattern is in what repeats. If several took off, look at what they have in common and what separates them from the rest. If only one took off, what separates it.

If you don't want to do the math by hand

I built a skill for Claude that does steps 2 to 4 on an account. You paste in the likes from the latest posts, yours or from an account in your niche, and it gives back the normal, each post's index, and which ones doubled the normal. It does that one thing.