The first hour after you publish is the whole job

Here is a pattern you can watch in almost any account. Someone spends two hours on a post. They rewrite the caption four times, crop the image twice, hunt for the right hashtags, hit publish at the time some article told them was optimal, and then close the app.

They open it again the next morning to check the numbers, and the numbers are disappointing. So they conclude the caption was wrong, or the time was wrong, or the algorithm hates them this week, and they spend another two hours on the next one.

The caption was probably fine. What went wrong happened after they closed the app.

Why the first hour carries so much weight

You do not need leaked internal documents to work this out. Every feed ranking system has the same problem to solve: a post appears, and the platform has no idea whether it is any good. Nothing about the text tells it. So it does the only sensible thing, which is to show it to a small slice of the audience and watch what they do.

That slice is small and the watching is quick. If people stop, read, reply, save or send it to someone, the platform widens the circle. If they scroll past, it does not. The post is not punished, it is simply not promoted, which amounts to the same thing.

This is why two posts of identical quality can end up an order of magnitude apart. One caught its first slice on a good day. The other did not, and there was nobody there to help it.

You cannot control the slice. You can control almost everything else that happens in that window.

The sixty-minute routine

None of this requires a desk. It is a phone routine, and once you have done it a dozen times it takes less attention than the caption did.

Minutes 0 to 5: seed the conversation

Post your own first comment. Not a hashtag dump, an actual sentence that gives people something to answer: the detail you had to cut, the question you were left with, the opinion you softened in the caption because it felt too sharp.

A comment thread with one comment in it invites a second. A thread with zero comments in it invites nothing. You are not gaming anything, you are refusing to leave an empty room.

Minutes 5 to 20: be present

Stay in the app. When a reply lands, answer it while the person is still there. A reply three minutes after the comment often produces a second comment from the same person. A reply the next morning produces nothing, because they have moved on and will never see it.

Answer with something that can be answered back. “Thanks!” closes a conversation. “Which size did you get?” keeps it open. You are not trying to be clever, you are trying to make the thread look alive to the next person who arrives.

Minutes 20 to 40: go where the people already are

Leave your own post. Go and comment, properly, on five or six accounts your audience follows. Not “great post 🔥”, an actual contribution: the thing you know that they did not say, the disagreement you can defend, the specific compliment.

Half the people who read a good comment look at who wrote it. That is the cheapest distribution available to a small account, and it costs nothing but attention.

Minutes 40 to 60: send it to three people

Not a broadcast. Three individual messages to three people who will genuinely care, with a reason attached: “this is the thing we argued about on Thursday.”

A share from a message is worth more than a share from a button, because it arrives with a recommendation attached. And on several platforms, a send is weighted more heavily than a like, because it is much harder to fake.

How to tell a slow post from a dead one

You do not need a benchmark from an industry report. You need your own last ten posts.

Write down, for each of them, how many views they had at roughly the one-hour mark and how many they ended with. Two things fall out of that list immediately.

The first is your own ratio. Most accounts settle somewhere predictable: a post that ends up at four or five times its first-hour number is normal, and a post that ends at ten times it did something unusual. Whatever your ratio is, it is yours, and it is far more useful than any published average.

The second is a threshold. Once you know your ratio, the number at one hour tells you where the post will land. That is the moment to decide whether to spend more energy on it or let it go.

What to do with a post that missed

The instinct is to delete it. Resist that. A post that underperformed is the only free research you get, and deleting it throws away the evidence.

Ask a narrower question than “why did this flop”. Ask where people stopped. If almost nobody saw it, the problem is the first line or the first frame, because that is all most people ever saw. If plenty saw it and nobody did anything, the problem is that it gave them nothing to do: no question, no opinion to agree with, no reason to send it to anyone.

Those are two completely different failures with two completely different fixes, and treating them as one thing called “bad post” is why people rewrite captions for months without improving.

Then reuse the idea rather than the post. A good idea that landed badly deserves a second attempt with a different opening, three weeks later, on a different day. Almost nobody remembers seeing the first one. That is not cynical, it is arithmetic: hardly anyone saw it, which was the problem.

The uncomfortable part

This routine takes an hour, several times a week, and it is not creative work. It is answering people. Most accounts that plateau have not run out of ideas, they have run out of willingness to do the unglamorous half.

It is also the half that compounds. Captions do not accumulate. Relationships with the forty people who reliably show up in your comments do, and those forty are what makes the first hour work at all.

Questions people ask about this

Does posting at the right time still matter?

Less than being available afterwards. The best time to post is a time when you can spend the next hour in the app. A perfectly timed post you abandon does worse than an average-timed post you tend.

Is replying to my own post gaming the algorithm?

No, and it would not work if it were. Adding one comment to your own post does not move a ranking system. It changes what the next human sees when they arrive, which is the actual mechanism.

What if I manage several accounts and cannot sit in all of them?

Then you triage instead of monitoring. Do the full hour on the one post that matters most this week, and for the rest, answer only what is a question or a complaint. Those two are the ones that cost you money when ignored.

How long before this shows up in the numbers?

Weeks, not days, and it shows up in comments before it shows up in followers. That order is normal and it is a good sign: the followers come from the people who saw the conversation, which has to exist first.