Most social media benchmarks come from surveys, or from someone else's clients. These are measurements. Across 7,945 posts on 336 accounts connected to Mydrop, in 111 workspaces and 9 platforms, this is what moved engagement and what did not, as of September 2026.
TLDR: Hashtag count barely matters (at most 5% either way). Captions under 80 characters underperform by 9%, and longer ones do not help. Carousels run 10% above the account's typical post, a link in the caption costs 7%, and posting between midnight and 6 am local time costs 17%. Weekday is noise.
How the numbers are measured
Every connected profile in Mydrop keeps a rolling 180-day history of its published posts, fetched from the platform itself: format, caption length, hashtag count, whether it carried a link, video duration, publish time, and the engagement rate the platform reports. The same data feeds the Patterns section of Performances for each account.
Three rules keep the pooled version honest:
- Medians, never averages. One viral post cannot move a bucket.
- Every post is scored against its own account's median. An index of 1.00 is a typical post for that account. 1.10 is 10% above. This is what lets a 200-follower bakery and a 2-million-follower brand sit in the same table, and what stops LinkedIn and TikTok rates from being averaged together.
- Buckets under 30 posts are withheld. If a cell is missing below, the sample was too thin.
Timing is measured in each workspace's own time zone. Mydrop's internal test workspaces are excluded. Posts are counted whether or not they were scheduled through Mydrop.
| Platform | Posts in the pool |
|---|---|
| 3,317 | |
| 2,777 | |
| 636 | |
| Google Business Profile | 397 |
| 347 | |
| TikTok | 219 |
| YouTube | 163 |
| Threads | 72 |
| Bluesky | 17 |
4,848 of the 7,945 posts carry an engagement rate; LinkedIn and Google Business Profile posts do not report one in a comparable way and sit out of the engagement tables.
Median engagement rate by platform
| Platform | Posts | Median | Middle half (25th to 75th percentile) |
|---|---|---|---|
| 2,901 | 3.08% | 1.06% to 6.71% | |
| TikTok | 193 | 2.98% | 1.77% to 6.55% |
| 1,630 | 1.11% | 0.36% to 3.14% | |
| 63 | 5.26% | 0.0% to 9.09% | |
| YouTube | 61 | 3.85% | 1.53% to 10.59% |
Read it as a range, not a target. A 2% Instagram post is normal. A 7% post is a good day. Pinterest and YouTube are thin samples and will move as the pool grows. The definition is the one in our engagement rate guide: interactions divided by reach.
Hashtag count: at most 5% either way
| Hashtags in the caption | Posts | Index |
|---|---|---|
| 0 | 1,624 | 0.96 (-4%) |
| 1 to 2 | 199 | 0.99 |
| 3 to 5 | 2,172 | 1.00 |
| 6 to 10 | 754 | 1.05 (+5%) |
| 11+ | 99 | 0.95 (-5%) |
The whole table lives inside a 10-point band. Posts with no hashtags run slightly under, 6 to 10 slightly over, and stacking 11 or more gives the small gain back. Hashtags are a filing label, not a reach lever, which is the argument in Do hashtags still work in 2026?, now with the numbers.
Caption length: only very short captions lose
| Caption length | Posts | Index |
|---|---|---|
| Under 80 characters | 702 | 0.91 (-9%) |
| 80 to 150 | 521 | 1.03 |
| 150 to 300 | 914 | 1.00 |
| 300 to 600 | 1,632 | 1.00 |
| 600+ | 1,079 | 1.01 |
Past 80 characters, length does nothing measurable. Write the caption the post needs and stop. The platform-by-platform cut-offs, where the feed hides the rest, are in the caption length guide.
Format: carousels win, links cost
| Format | Posts | Index |
|---|---|---|
| Carousel | 749 | 1.10 (+10%) |
| Photo | 2,102 | 1.00 |
| Video | 1,866 | 0.99 |
| Link in the caption | Posts | Index |
|---|---|---|
| No link | 4,582 | 1.00 |
| With link | 266 | 0.93 (-7%) |
Video length, on the 222 videos that report a duration: 30 to 60 seconds 1.06, 60 to 180 seconds 1.03, under 15 seconds 1.00, 15 to 30 seconds 0.94. Thin buckets; treat the ordering as a hint, not a rule.
Timing: the weekday is noise, the night is not
Measured in each account's local time.
| Weekday | Posts | Index |
|---|---|---|
| Monday | 792 | 0.99 |
| Tuesday | 696 | 1.00 |
| Wednesday | 839 | 1.00 |
| Thursday | 669 | 1.00 |
| Friday | 847 | 1.00 |
| Saturday | 518 | 1.02 |
| Sunday | 487 | 0.99 |
| Time of day (local) | Posts | Index |
|---|---|---|
| Morning, 6 am to noon | 1,684 | 1.00 |
| Afternoon, noon to 6 pm | 2,162 | 1.00 |
| Evening, 6 pm to midnight | 863 | 1.00 |
| Night, midnight to 6 am | 139 | 0.83 (-17%) |
Every weekday sits within 2% of the median. The only timing effect worth acting on is the overnight slot: -17% across platforms and -36% on Instagram (n=90). If your scheduler fills empty slots at 2 am, that is the fix. Your own account's grid, by weekday and hour, is what Performances shows once it has 8 posts; that beats any pooled table, including this one. The general method is in how to find your best posting times.
Posting cadence: read this one carefully
| Posts per week | Accounts | Median engagement rate |
|---|---|---|
| Under 1 | 16 | 7.03% |
| 1 to 3 | 80 | 3.08% |
| 3 to 7 | 55 | 1.46% |
| 7 or more | 39 | 0.68% |
This is a correlation, not a cause. Small accounts post less and carry higher rates because a few loyal followers make up most of their audience. It does not say that posting less raises engagement. It does say that if your rate fell as you scaled up posting, you are normal.
Publishing reliability and approval speed
Two operational numbers from the product itself, separate from the engagement pool:
- Of 726 posts published through Mydrop in the 30 days to 16 September 2026, 98.5% went live at the first attempt. The most common warning was a YouTube custom thumbnail rejected because the channel was not phone-verified. The most common hard failure was an expired platform connection.
- On the 41 posts that went through a client approval with both timestamps recorded, the median time from submission to approval was 6 hours, and 66% were approved within 24 hours. Small sample, quoted as is.
How to use this on your own accounts
- Stop optimising hashtag count and weekday. The data says they are noise.
- Cut captions under 80 characters unless the image carries the whole message.
- Turn a photo post into a carousel when you have three or more images. That is the one format lever with a measurable lift.
- Move links to the first comment or your link in bio page when reach matters more than the click.
- Check your scheduler for overnight slots.
Then open your own patterns. Every account connected to Mydrop gets this same analysis on its own posts, by weekday, hour, format, length and hashtag count, once it has 8 posts of history. Your numbers replace ours.
Refresh schedule
This page is regenerated from the live pool. Next refresh: December 2026. Sample sizes in the tables are the ones at the time of the last run, 2026-09-17.
Google review
Trustpilot review
Product Hunt comment
