Posting more often drops per-post engagement rate about 4×. It also raises total weekly engagement 5.4×. Both are true, only one of them matters, and the industry has been quoting the wrong number for years.
TL;DR
The two charts that disagree
Every existing posting-frequency study reports the chart on the left and stops there. The chart on the right is the same accounts, same engagements, just divided by week instead of by post. It tells the opposite story.
Per-post ER = (likes + comments) / followers, median across the account's last 12 non-pinned posts. Weekly ER = total engagements in the sample window, normalized to per-week and per-follower. Sample of 2,400 accounts.
The consensus, and what it gets wrong
Buffer's 27M-post engagement study, Sprout Social's benchmark reports, Hootsuite's algorithm guide, Tailwind's daily-posting study, the Zoomsphere “Frequency Formula” report - every credible posting-frequency analysis published in the last three years measures per-post engagement rate and concludes that posting more frequently “dilutes” engagement. They are not wrong. The math is correct.
The problem is that per-post engagement rate is a ratio with the number of posts in the denominator. If you keep total engagements constant and double your posts, your per-post ER halves automatically. It would halve if Instagram's algorithm did nothing at all. Reporting that a metric whose denominator is “number of posts” goes down when you post more is mathematically uninformative.
The metric that actually answers “should I post more?” is total weekly engagement - likes + comments delivered to your audience per unit of time. We computed this on the same 2,400 accounts. It rises with cadence in every follower size cohort we have meaningful sample for. The per-post decline is real, but it is not steep enough to outweigh the volume increase.
The advice that gets quoted in “Instagram posting frequency” articles is technically correct and practically backwards. If you ran a clothing brand and your CMO told you to ship fewer ad campaigns per quarter because your “engagement per campaign” was dropping, you would fire them. The same logic applies to posts.
Controlling for follower size
The obvious objection to the headline finding: bigger accounts post more and have lower per-post ER, so the correlation might just reflect account size. To rule that out, we split by follower size and re-ran. Within every cohort, total weekly engagement still rises with cadence - and the effect is sharpest where it matters most.
| Follower size | <1/wk | 1-2/wk | 2-4/wk | 4-7/wk | 7-14/wk |
|---|---|---|---|---|---|
| 1k-10k | 1.71% n=210 | 3.82% n=112 | 7.28% n=141 | 13.89% n=75 | 20.01% n=38 |
| 10k-100k | 1.76% n=229 | 2.00% n=167 | 5.02% n=249 | 4.18% n=160 | 8.75% n=121 |
| 100k-1M | 1.59% n=150 | 4.47% n=91 | 3.50% n=192 | 4.22% n=164 | 9.10% n=108 |
| 1M+ | 2.19% n=52 | 7.63% n=21 | 5.15% n=49 | 8.45% n=21 | 5.31% n=30 |
Darker cells = higher total weekly engagement. The 1k-10k row shows the strongest cadence effect - small accounts posting daily generate ~12× the weekly engagement of those posting weekly.
The headline cohort
1k-10k followers - the cohort where most creators and small brands actually live. Total weekly engagement rises from 1.71% (<1/wk) to 20.01% (7-14/wk). That is a 12× multiplier. The “post less” advice is the most damaging here, because nano accounts have the most room to grow and the lowest cost to publish more.
Where it gets noisy
1M+ accounts have small sample sizes per cell (n=21 to n=52) and the cadence curve flattens. The trend still points up but cohort variance is high. At that scale, content quality and format mix probably matter more than raw cadence - there is no obvious diminishing-returns cliff in our data, but we would not stake a claim on the 1M+ row alone.
Niche variation
The benchmarks reported by major social tools blend creators with retail brands with media accounts. The result is a single industry average that misleads everyone. Here is what the same metrics look like split by niche.
| Niche | n | Median posts/wk | Per-post ER | Weekly ER |
|---|---|---|---|---|
| Creator (artists, digital creators, reel creators) | 131 | 1.60 | 1.44% | 3.63% |
| Business (coaches, entrepreneurs, consultants) | 40 | 2.27 | 0.74% | 1.67% |
| Design (furniture, architecture, interiors) | 32 | 2.05 | 0.16% | 0.53% |
| E-commerce (retail, product/service brands) | 30 | 2.49 | 0.23% | 0.60% |
Creators get ~9× the per-post engagement that design accounts get, at a lower posting cadence (1.60/wk vs 2.05/wk). A creator with a 1.0% per-post ER is below their niche median. An interior-design account with the same 1.0% is doing extraordinary work. The single-number benchmarks pretend these are comparable. They are not.
Caveat: the within-niche cadence curve is only supportable for creators (n=131). Business, design, and e-commerce samples are large enough for descriptive medians but too thin for a within-niche cadence breakdown. We will publish niche-specific cadence studies once those samples grow.
What this means for you
It is a ratio where the number of posts is in the denominator, and it is the metric every other tool reports because it produces a tidy headline number. If you care about reach, growth, or sales, what matters is engagements per week per follower - total volume, not per-post rate.
The 12× weekly engagement multiplier in the 1k-10k cohort is the clearest finding in this study. The cost of a low per-post ER at this size is essentially zero - nobody is benchmarking your individual posts. The cost of posting less is real and compounds. Cap by content quality, not by cadence.
A “good” engagement rate for a creator is a terrible one for a design account, and vice versa. The real benchmark is your peer set: the accounts your audience already follows. That is a different calculation per account, which is why no single-number benchmark can be right for you.
Apply this to your own account
Competitor Map finds the accounts that are actually in your niche and shows you their best performing posts and median engagement. Unlike other tools, it works by analyzing your audience and using AI to find similar accounts, instead of relying on keyword searches that surface unrelated handles.
Methodology
Sample
2,400 public Instagram accounts with 10+ recent posts stored, follower counts between 1k and 10M. Accounts were discovered through SocialToolsAI's audience-overlap scans across a range of niches - they were not hand-picked.
Posts per account
The most recent 12 non-pinned posts. Each post has a stored timestamp, like count, and comment count.
Posting frequency
Computed as (number of intervals between posts in the sample) × 7 / (days between oldest and newest post in the sample). Accounts where the 12 posts span less than 7 days were excluded to avoid noise from recent posting bursts.
Per-post engagement rate
For each post: (likes + comments) / follower count. Account-level value is the median across the 12 posts. Median instead of mean to suppress the effect of one viral post.
Weekly engagement rate
Total likes + comments across all 12 posts, divided by the span in weeks, divided by follower count. This gives engagements delivered to the audience per week, normalized to account size.
What this study cannot prove
This is observational data, not an experiment. We can show that accounts which post more tend to have higher weekly engagement. We cannot prove that your account would gain weekly engagement by posting more - accounts that post less may self-select for higher-effort content, audiences may differ, and reverse causality (accounts that grow fast post more because they have more to talk about) is plausible. Treat the macro finding as a hypothesis to test on your own account, not a deterministic rule.
Questions people ask
Related
Map every account sharing audience with any public handle. The starting point for niche-aware benchmarking.
Identify the real competitive set your followers also pay attention to. Includes median engagement on every result.
Once you know your cadence, what to actually post. Reel ideas generated from the top posts in your niche.