How often to post on Instagram, and how to keep it up

There is no number that is right for everybody, because the right number depends on how long a post takes you and nobody publishing a number knows that. What the evidence supports is narrower and more useful: posting more helps, posting nothing hurts, and the outcome over a year is decided by whether you are still going in March. So the question worth answering is not which cadence is best. It is which cadence you can survive.

Here is what the research does and does not show, how to find your own number with a clock, and what to do in the week you cannot manage it.

Why nobody can give you a number

The right number depends on how long a post takes you, and that varies enormously. Somebody photographing the counter and writing two lines is at ten minutes. Somebody building a twelve slide carousel from a blank page is at ninety. Give both of them the same advice and you have described two completely different weeks. Any published number that does not know your minutes is guessing at the variable that decides everything.

It also depends on how much arrives on its own. A kitchen that changes what it makes every day has material walking in the door. Work where the same job is done well, over and over, has to go and find the material, and finding it is a separate task with its own cost. Both are normal. They are not the same weekly commitment.

And a published number is an average of other people's weeks. Averages are made mostly of accounts unlike yours, and the ones at the top of the range are rarely being run by one person between other jobs. What transfers between them and you is the shape of the finding, never the number itself.

What the research does and does not show

The best available evidence comes from Buffer, and it should be read with the conflict in view: Buffer sells scheduling software, so "post more, and post regularly" is the conclusion its software benefits from. That is a reason to check the method rather than a reason to dismiss the finding, and on one of its two studies the method is unusually good.

The frequency study covers 2.1 million posts across about 102,000 Instagram accounts, published by Buffer and read by us in September 2026. The design is the interesting part. Rather than comparing accounts with each other, it compares each account against its own baseline, so a large account's built-in advantage cannot be mistaken for an effect of posting. Growth rose at every step up in frequency.

Read that column as a ranking rather than as a forecast. It is a rate measured over Buffer's own window on Buffer's own users, and your account is not in it. What travels is the direction, and that it holds at every step rather than only at the top.

Two more things from the same analysis are worth having. Accounts posting three to five times a week got about 12% more reach per post than those posting once or twice, so at that end more posting was not diluting anything. And the band for periods with no posts at all is the only one that comes out below zero rather than merely low. Not posting was not neutral.

What the study cannot tell you is what happens after you stop, because it is measuring accounts during a period in which they were posting.

The consistency study is the one people quote and the one to read more carefully. In a second piece published in January 2025, Buffer took over 100,000 of its users across ten platforms and counted how many of 26 weeks each one posted in. Accounts that posted in 20 or more of those weeks got roughly five times the engagement per post of accounts that posted in four weeks or fewer, with the middle group at about 3.4 times.

That design is weaker, and the difference matters. It compares different people with each other rather than each account against itself. The people who managed 20 weeks out of 26 are not a random sample of the people who managed four. They are more likely to be the ones whose posting was already working, and something working is the single best predictor of somebody keeping it up. So the finding shows that consistency and results travel together. It cannot tell you which way the arrow points.

What survives all of that is smaller than the headlines and more useful than either. More is better if you can hold it, holding it is the binding constraint, and nothing in any of this says the number has to be large.

Finding a number you can hold

Ten minutes and a clock, once.

Time yourself making one post honestly, end to end: deciding what to say, writing it, making the picture, the caption, the upload. Not the fastest one you ever made. A typical one, on a day when you were not in the mood. Then multiply it out.

Now pick the row you would still agree to on a Wednesday in November when something has gone wrong at work. Not the row you would like to be. The distance between those two rows is where nearly every abandoned account lives, and it is chosen in a single optimistic afternoon.

The middle columns are the lever people forget. Halving your minutes per post moves you a row or two up the table without asking anything more of you, and most of those minutes are spent deciding what to say rather than making anything, which is why writing a month of ideas in one sitting is worth more than any editing shortcut. Working in a fixed order is the other half: the order that saves the most work is the version for a carousel, and the visual decisions get faster too once the six that matter on a slide are settled once instead of every time.

Then set the rate on a one year view rather than a one month view. Two a week is 104 posts by this time next year, and 104 posts is a body of work that people find, search through and scroll back into. Almost nobody believes that in week three, which is exactly why so few people ever get to see it.

The week you cannot

Drop down, do not disappear

Posting less and stopping are not the same thing at different sizes. They are different events.

People unfollow at roughly the same rate whether or not you posted. They tidy their feed, lose interest, change jobs, leave the platform. None of that pauses when you do, and in a quiet month nothing is arriving to offset it, so a gap does not merely fail to add. It takes away. That is the plainest reading of Buffer's no-posting band coming out negative rather than flat, and you can watch the same thing happen on any account you follow that went quiet last spring.

The second cost is entirely in your head and it is the larger one. The longer you have been gone, the more the return post feels like it has to account for the absence, so it has to be a good one, so you do not make it on Tuesday. Six weeks becomes six months by exactly that route, and it starts with one skipped week that felt like nothing at the time.

So the move in a bad week is not to hold the standard and miss. It is to lower the standard and land. One post you are not especially proud of keeps the rhythm, and the rhythm is the thing with value in it.

Keep two spares

The cheapest insurance available, and almost nobody has it.

Two finished posts, sitting ready, with nothing time bound in them. No date, no offer, no event. The answer to the question you are asked most often, or the thing you find too obvious to mention, or a correction of something people get wrong. Those keep indefinitely and they are useful in any week of any year.

Spend one in the week that goes wrong, and refill it the week after. Two rather than ten, deliberately: a reserve of ten stops being a reserve and becomes a schedule, and a schedule written in August gets posted in November by somebody who has stopped paying attention to what changed in between.

If the spares are gone and the week is gone too, the smallest honest post is one slide, one sentence and a photograph you already have. What to post when nothing new has happened is a list of those.

Mistakes

• Announcing the break. "Sorry it has been quiet" is a post about you, made of nothing, and it informs everybody who had not noticed that you were gone. Post the next real thing instead and say nothing about the gap.
• Choosing the cadence on a good day. Deciding is itself an enthusiastic act, so the decision always gets made in an optimistic hour. Pick the number as though you were choosing it in your worst month, because that is the month it has to survive.
• Treating one missed post as a failed run. A missed post is a missed post. Deciding it proves the whole thing is not working is how a slow week turns into a dead account, and it is a much bigger decision than the one you actually made.
• Confusing plain with empty. A plain post that tells somebody something is fine forever. A post that exists only so that something was posted teaches people to scroll past you, which is expensive and slow to undo. Lower the polish, never the point.
• Judging the week by the wrong number. Most dashboards show you one engagement total, dominated by likes, which is the cheapest action anybody can take. People quit on that number. Look at saves and sends instead, since those are the two with a second life.
• Making the return post an event. It does not have to be. It just has to be Tuesday's.

What differs by platform

The mechanics travel. Buffer's consistency finding ran in the same direction across all ten platforms it covered, and the reason it would is not mysterious: everywhere, an account that appears regularly is one people remember, and an account they have not seen since spring is one they have quietly stopped expecting.

What genuinely changes is how long a single post keeps working, and that changes what your number should be. Where a post is finished within a day, you need more of them, because volume is the only way to be present at all. Where a post keeps being found for months, through search or through a recommendation surface, fewer and better wins, because each one is still earning in February. You do not have to guess which you are in. Look at whether anything you posted a month ago got seen last week: if it did, slow down and make them properly, and if nothing does, the cadence is carrying the work.

The other difference is what a gap looks like from outside. On a feed, going quiet is invisible until somebody wonders whether you are still going, and by then they have wondered. Nobody sends a message to check.

So time one post this week and multiply it out. Pick the row below the one you want, not above it. Then make two spares and put them somewhere you will actually find them in the week everything goes wrong, because that week is coming and it is the one that decides the year.