How-To Guide
How often should a brand post on Instagram: Reels, carousels and Stories?
Dash Social, Buffer and Metricool disagree on Instagram posting frequency because they measure different things. A 3 to 5 posts a week routine by format.
By the Superdirector editorial team. Vendor studies and Instagram pages read November 7, 2026; reviewed November 11, 2026.
At a glance
How often should a brand post on Instagram: Reels, carousels and Stories?
Most brands should post 3 to 5 times a week to the Instagram feed, split between carousels and Reels, and use Stories on most days. Beyond that the big datasets disagree because they measure different things: Dash Social found the best engagement rate per post at 2 to 3 posts a week, Buffer found reach per post and follower growth keep rising up to 10 or more, and Metricool and HypeAuditor found average yearly follower growth of 19% at 2 to 3 posts and 79% at 10 or more. Buffer and Metricool both show accounts that post once a week or less stalling or shrinking. Pick a number you can keep for eight weeks, then test one step up against your own median.
- Feed posts (Reels, carousels and images together): 3 to 5 a week, with 1 to 2 as the minimum before Buffer saw stagnation.
- Reels: 1 to 3 a week had the best engagement rate in Dash Social's data; Stories completed best at 30 or more a week.
- Every figure is an average or a correlation from other accounts; test one step up against your own median.
Each study measures something different, so the shapes differ. Original diagram; figures from Buffer, Metricool with HypeAuditor and Dash Social.
Make it work for your brand
Find it. Remake it. Post it. Superdirector turns the viral formats in your niche into your brand's carousels, X posts and video scripts, and keeps the calendar full.
How often a brand should post on Instagram, by format
Post 3 to 5 times a week to the feed and add Stories on most days. That range is where the largest published studies overlap. Buffer's analysis of 2.1 million posts from 102,000 accounts found the biggest single step in reach per post, about 12%, when an account moved from 1 or 2 posts a week to 3 to 5. Hopper HQ and Kontentino recommend the same 3 to 5 feed posts, though neither shows a test behind the count.
The table gives a starting point for each format and what the number is based on. Treat every row as a place to begin, not a target: Dash Social, whose data appears below, calls its own figures "averages, not requirements."
| Format | Starting cadence | What the evidence says | What was measured |
|---|---|---|---|
| Feed posts (Reels, carousels and images together) | 3 to 5 a week | Buffer: about 12% more reach per post at 3 to 5 a week than at 1 to 2. Dash Social: best engagement rate at 2 to 3 a week. | Reach per post (Buffer); engagement rate per post (Dash) |
| Reels | 1 to 3 a week, inside that total | Dash Social: highest Reels engagement at 1 to 3 a week. Kontentino recommends 4 to 7 but cites no test of it. | Engagement rate per post |
| Carousels | 2 to 3 a week, inside that total (our split) | Socialinsider: carousels averaged a 0.55% engagement rate against 0.52% for Reels and 0.37% for images in 2025, while the average brand posted 5 carousels and 8 Reels a month. | Engagement rate by followers |
| Stories | Most days | Dash Social: highest completion rate at 30 or more Stories a week, about 4 or 5 a day. Each frame counts as a Story. | Completion rate |
| Minimum | 1 to 2 feed posts a week | Buffer: posting any less risks stagnation or a drop in followers. | Follower growth against the account's own baseline |
Two limits on these figures. They are averages across other people's accounts, and most of them are correlations. Buffer's method comes closest to a fair test because it compares each account with itself, week by week. The next section shows why the studies still point in different directions.
Why three big datasets give three different answers
Each study answers a different question on a different set of accounts. One asks which posting rate gives the best engagement per post. Another asks which gives the most reach or follower growth. A third only counts how often accounts post. Put the outcome next to the number and most of the disagreement goes away.
| Study | Accounts and posts | Outcome it measures | What it found | Limit |
|---|---|---|---|---|
| Dash Social, data July 2025 to June 2026 | About 3,000 brands, 1.5 million feed posts and 1.1 million Stories | Average engagement rate per post; Story completion rate | 2 to 3 feed posts a week scored highest; 1 to 3 Reels; 30 or more Stories | A per-post rate says nothing about total reach across the week |
| Buffer, published August 2025 | 2.1 million feed posts from 102,000 accounts | Follower growth rate and reach per post, each account against its own baseline | Reach per post about 12%, 18% and 24% higher at 3 to 5, 6 to 9 and 10 or more posts a week than at 1 to 2 | Buffer's own users; feed posts only; the first step up is the biggest |
| Metricool and HypeAuditor, January to June 2025, as summarised by Netinfluencer | 700 million posts from 28 million influencer accounts | Average yearly follower growth | 19% at 2 to 3 posts a week, 79% at 10 or more; most accounts post once a week or less | Influencer accounts, compared with each other; a correlation, not a controlled test |
| Later, January to October 2023 | 19 million feed posts and 32,000 Stories | How often accounts post | Micro accounts (10K to 100K followers) averaged 3 feed posts a week; mid accounts (100K to 500K) averaged 15 Stories a week | Describes behaviour; says nothing about results |
| Socialinsider, 2025 | 67.5 million posts from 579,000 brand accounts | Engagement rate by followers; format mix | The average brand published 8 Reels, 7 images and 5 carousels a month | Describes behaviour and engagement, organic posts only |
Read together, the studies agree on three things. Buffer and Metricool both show accounts that go quiet losing ground. In Buffer's data the step from 1 or 2 posts to 3 to 5 gives the largest gain in reach per post. And Buffer finds that more than 5 a week still adds reach and growth, with each later step smaller than the first. Whether more than 5 is worth it depends on what you can make well, and no study here can answer that for your account.
Set a floor you can keep for eight weeks
Start at 3 feed posts a week, or 2 if that is all you can make well. Buffer's data shows why the floor matters more than the ceiling: weeks with no posts under-performed an account's own baseline, and posting any less than 1 to 2 times a week risked stagnation or a drop in followers. Netinfluencer's summary of the Metricool and HypeAuditor study points the same way: 93.5% of accounts post once a week or less and lose 2% of their followers a year on average. A steady 3 beats a burst of 7 followed by a silent week.
Before you pick the number, write down the posts you could actually finish in the next 14 days: topic, format and day. Halve that list to get a weekly count, and start there even if it is below 3. A count you have already planned is one you can keep. The table turns a few common situations into a starting point. The splits are ours, built from the evidence above, not a study's result.
| Your situation | Start with | Why this split |
|---|---|---|
| You post alone and do not film video yet | 2 to 3 carousels a week, Stories on 3 or 4 days | Carousels can be made from photos and text you already have, and 2 to 3 sits in Dash Social's best engagement band |
| You can film one Reel a week | 2 carousels and 1 Reel a week, Stories most days | Reaches the 3-post floor and puts one post in front of non-followers through Reels |
| Someone helps you a few hours a week | 3 carousels and 2 Reels a week, Stories most days | Puts you at 5, the top of the 3 to 5 range, where Buffer still measured reach per post rising |
| The account has been quiet for a month or more | 2 feed posts a week for two weeks, then your floor | Rebuilds the habit first; in Buffer's data, weeks with no posts did worse than the account's own baseline |
Instagram publishes no ideal number in its ranking explainer. It describes the signals, the viewer's history with you, information about the post and information about you, and says Feed ranking tries to avoid showing too many posts from the same person in a row. That is the only official hint at spacing: five posts across a week are not five posts in an hour.
The nearest thing to an official cadence is inside the app. When Meta launched Instagram's Best practices hub, it said the Creation category gives insights on how often to post, and that each category includes tips personalised to the account. That section sits in the professional dashboard and we could not read it, so open yours and compare it with the numbers above.
One number often credited to Instagram is old. Inmagic traces the "two feed posts a week, two Stories a day" line to a June 2021 Creator Week Q&A, and we found no Instagram page for it. For tools that publish through Instagram's API, the developer documentation sets a rate limit of 100 API-published posts in a 24-hour moving period, with a carousel counting as one, though the carousel section of the same page says accounts are limited to 50 published posts in 24 hours. Either way, no human routine comes near it.
Give each format a job
Split your weekly count by what each format is good at. Instagram says each part of the app, Feed, Stories, Explore and Reels, uses its own algorithm, so a post count that works for one format does not carry over to another.
| Format | Who sees it | What Instagram says it predicts | What the datasets add |
|---|---|---|---|
| Reels | Mostly people who do not follow you: Instagram says the majority of what people see in Reels is from accounts they do not follow | Reshare, watching all the way through, like, and visiting the audio page | Metricool and HypeAuditor, via Netinfluencer: Reels reach the most people for accounts under 1,000 followers (median 134 views against 56 for carousels and 23 for images) |
| Carousels | Followers in Feed, and non-followers in Explore, where Instagram says likes, saves and shares are the most important predicted actions | A few seconds spent, comment, like, share, profile tap in Feed | Socialinsider: carousels earned the most saves for brands above 10,000 followers, up to 98 per post at 100,000 to 1 million |
| Stories | People who already follow you; Instagram says Stories come from accounts the viewer follows, ranked on viewing history, engagement history and closeness | Tap into a Story, reply, move on to the next | Dash Social: best completion at 30 or more a week. Buffer left Stories out because they are not typically shown to non-followers |
Two cautions follow from the table. First, Stories are not a substitute for feed posts. Instagram's head Adam Mosseri said in August 2024 that feed posts usually get a lot more reach than Stories, because there is far more Stories inventory competing for attention. Second, do not raise your count by reposting the same Reel. Instagram's ranking page lists reels that have already been posted on Instagram among those it makes less visible in recommendations, so a new angle on the same idea counts and a repeat does not.
If you cannot film Reels every week, a steady run of carousels is a reasonable base. Socialinsider's engagement rate for carousels was at or above Reels in 2025, and a carousel can be made from text and images you already have. Add Reels when you can make them well.
Reels reach people who do not follow you, carousels are ranked on saves for followers, and Stories reach regulars. Original diagram; audience signals from Instagram's ranking explainer, splits from this guide.
Run a four-week test that changes only the count
Run your floor for two weeks, then one step higher for two weeks, with the same formats, topics and posting times. Changing the count and the content together tells you nothing about the count. The two example weeks below are our split, not a study's finding. Stories run on the same five days in both weeks, so the only thing that changes is the number of feed posts.
| Day | 3 feed posts a week | 5 feed posts a week | Stories (same in both weeks) |
|---|---|---|---|
| Monday | Carousel | Carousel | Stories |
| Tuesday | No feed post | Reel | Stories |
| Wednesday | Reel | Carousel | Stories |
| Thursday | No feed post | No feed post | Stories |
| Friday | Carousel | Reel | Stories |
| Saturday | No feed post | Carousel | No Stories |
| Sunday | No feed post | No feed post | No Stories |
Socialinsider found brands lean towards Reels, 8 a month against 5 carousels, so if your mix is the reverse, expect your numbers to differ from the averages above and judge them against yourself.
A worked example: one shop's four-week test
Here is the test run once, with made-up numbers so the arithmetic is easy to follow. The shop is fictional: a small candle brand with about 2,000 followers, run by its founder, posting 3 times a week. Its weekly median reach per post over the four weeks before the test was 380, 410, 395 and 420.
Each week the founder writes down that week's median reach per post and divides it by the median of the four weeks before it. In week 1 that baseline is the median of 380, 410, 395 and 420, which is 403. In week 2 the window moves on by one week, so it becomes the median of 410, 395, 420 and 410, which is 410.
| Week | Feed posts | Median reach per post | Median of previous 4 weeks | Ratio | Saves plus shares per post |
|---|---|---|---|---|---|
| 1 (floor) | 3 | 410 | 403 | 1.02 | 6 |
| 2 (floor) | 3 | 390 | 410 | 0.95 | 5 |
| 3 (step up) | 5 | 430 | 403 | 1.07 | 7 |
| 4 (step up) | 5 | 415 | 415 | 1.00 | 6 |
Week 2 dipped to 0.95 at the same count, which shows how much an ordinary week moves on its own. That is why the test runs two weeks at each level and why one bad week is not a verdict. In weeks 3 and 4 the ratio stayed at or above 1.00 and saves plus shares did not drop, and the founder judged all ten posts good enough to publish. By the rule in the next sections, the shop keeps 5 a week. Had week 4 come in at 0.85 with two rushed posts, it would go back to 3.
Your own numbers will be different, and a small account can swing more from week to week than this one. The point of the example is the method: one change, your own baseline, two weeks per level, and a decision written down before you look.
The ratio held at or above baseline in both step-up weeks, so the shop keeps five a week. Original diagram; made-up numbers from the worked example above, not a real account.
Compare each week with your own median
Read three numbers in Instagram Insights each week: median reach per post, follows gained, and saves plus shares per post. They line up with what Instagram says it predicts, since Reels are ranked on resharing and watch-through, and Explore on saves and shares. Compare each week with your own previous four weeks, the same idea as Buffer's within-account comparison, and not with a vendor's benchmark built from other accounts.
The cadence test log has columns for planned and published posts, the format split, Stories, median reach, follows and the ratio against your previous four weeks. Use the median, not the average, so one post that travels does not hide the rest.
Keep the higher cadence only if quality held
Keep the step up if median reach per post did not fall and each post still met your own quality bar. If reach per post dropped, go back down. That is our rule of thumb, not a published threshold, and Buffer's own data says the extra reach per post is smaller after the first step, so the higher count has to earn its extra work.
For the definition of cadence and how to count a carousel, see posting cadence. For the Reel side, the Reels growth guide covers the signals Instagram names, and the content calendar guide shows how to keep the schedule alive past week three.
Where Superdirector fits
Superdirector publishes this guide, so weigh what follows with that in mind. You do not need any paid tool for the routine above: Instagram Insights and the free test log are enough to pick a floor, run the test and keep the result.
What Superdirector sells is help with the making and posting. Its $199 a month plan is set up to run one brand account for founders, app owners and DTC brands: it watches the accounts and topics you name, makes carousels and X posts in your brand, fills the month's calendar and publishes eligible Instagram carousels to the connected account under the account-level standing posting permission you set. The $29 a month plan is for doing the work yourself from Claude or ChatGPT, with ideas, scripts and drafts and no scheduled posting. The site's headline is "Autonomous social media growth.*" with the note "* AI video and full autopilot are coming soon," so Reels still need filming on your side, and the account-level permission stays yours. See what runs today.