Every time you post, the same thing happens. Your reel is shown to a small group of people. What they do decides whether it is shown to a larger one. That loop repeats, or it stops.

That is it. There is no queue you have been put in and no rule about posting at eight in the evening. There is a test, and it asks one question: did the people who saw this behave as though it was worth their time?

The signals, in the order they matter

  • Watch time and rewatches. How much of it they saw. A 15 second reel watched twice beats a 60 second reel abandoned at eight.
  • Sends. The strongest signal on the platform. One person sending your reel to a friend is worth a great many likes, because it costs them something to do it.
  • Saves. A promise to come back. Recipes, checklists, explanations and anything with steps earn these.
  • Comments. Real ones. A thread of replies counts for far more than a row of emojis.
  • Likes. Last, and by a distance. The cheapest action on the app, and priced accordingly.
  • Follows from that video. The clearest statement that somebody wants more of this.
Likes tell you a video was pleasant. Sends tell you it was useful.

What quietly costs you

Nothing here is a punishment. It is all the same test, failed in different ways.

  • A slow opening. Most videos die in the first three seconds, and it is nearly always a greeting.
  • Another platform's watermark. Instagram would rather not advertise for anyone else.
  • Recycled clips with borrowed audio. The platform can see it has shown this to people before.
  • Length with nothing in it. A long video is not rewarded for being long. It is measured on how much of it people finished.
  • A subject that changes every week. The system is trying to work out who to show you to. Give it something consistent to learn.

Reach is the number you should look at least

Reach is loud, it moves every day, and it mostly tells you how well one opening line worked. It says little about whether the page is growing.

Open your own insights and read them in this order instead. Where people stopped watching, which is the single most useful graph you have. How many saved it and how many sent it. How many followed from the video itself.

A reel with lower reach and twice the sends is the better video, every time. Make more of that one.

Posting daily is not the answer

Nine good videos beat thirty rushed ones, and thirty rushed ones will teach you less, because you will not have looked properly at any of them.

The habit worth building is not daily posting. It is looking back. Once a week, open the last four videos, find the two that held people longest, and ask what those two had in common. Usually it is the opening. Sometimes it is the subject. Occasionally it is that you were enjoying yourself, and it shows.

The full algorithm breakdown, with successful accounts taken apart on screen, is day five.