Automation makes your calendar look healthy even when it is not. Posts go out on time, every time, and nobody checks whether they work. This scorecard takes 15 minutes a month and tells you which scheduled posts to repeat, rewrite or drop.
TLDR: Take your last 30 scheduled posts, divide each one's engagement rate by your account median, and sort them into four buckets: repeat, rewrite, test again, retire. In Mydrop's data, the middle half of Instagram posts ran between 1.1% and 6.7%, so judge yourself against your own median, not a round number.
Mydrop data, September 2026: in the 30 days to 16 September 2026, 98.5% of posts published through Mydrop 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. Method in the Mydrop social media benchmarks.
Build the scorecard in six columns
Open a spreadsheet and list your last 30 posts, one per row. You need six columns:
| Column | What goes in it |
|---|---|
| Post and date | Link or title, publish date |
| Slot type | Evergreen, campaign, repost, reply to news |
| Format | Photo, video, carousel, text |
| Engagement rate | Engagements divided by reach (or followers if reach is missing) |
| Ratio to median | The rate divided by your median of the 30 rows |
| Verdict | Repeat, rewrite, test again, retire |
Find the median first, then fill the ratio column. The median matters more than the average because one viral post drags an average up and hides the 25 ordinary ones.
Mark outliers with two thresholds
Use fixed cut-offs so you do not argue with yourself row by row.
- 1.5x your median or above: an outlier on the high side. Find out why before you repeat it.
- 0.8x to 1.5x: normal range. Leave it alone.
- Below 0.8x: weak. Below 0.5x is a real miss.
In Mydrop's September 2026 data, a 2% Instagram post is normal and a 7% post is a good day. Accounts differ a lot, so your own thresholds beat any published number.
Read what the automated slots are doing
Now group the rows by slot type and take the median ratio of each group. This is where automation leaks show up.
- Recurring slots that sit under 0.8x (the "Monday tip" nobody asked for) are running on habit. Rewrite the format or retire the slot.
- Reposts and recycled posts that match or beat their original are worth scheduling again. If they fall under 0.5x, your audience has seen them.
- One-off posts that beat 1.5x show you what the audience wants. Turn the idea into a recurring slot and measure it for another month.
One more pattern to check is time. In our data, posts published between midnight and 6 am local time ran 36% below median on Instagram. If your low rows cluster at night, the leak is the slot time, not the content.
Give every verdict a next action
Do not leave the verdict column as a label. Attach an action and an owner.
- Repeat: schedule two more in the same format within two weeks.
- Rewrite: same topic, new hook or format. Carousels ran 10% above the typical post in our data, so try one.
- Test again: posts between 0.8x and 1.0x with a weak hook. Change one thing, post once more.
- Retire: delete the recurring slot from the calendar.
Retiring matters. A calendar full of 0.6x posts trains you to treat output as success.
Pull scheduled-post numbers into one sheet
The hard part of this audit is gathering the data, not scoring it. If you copy engagement numbers from each native app, it takes an afternoon per account. Mydrop's performance view shows each post against that account's own median, and scheduled reports can send the monthly summary to a client or a manager without anyone exporting a CSV.
Teams running several brands benefit most, because each brand gets its own median instead of being compared to the loudest account in the group.
Run it this week on your noisiest account
Pick the account that posts the most and has the weakest feel. Score its last 30 posts in the table above, retire the bottom three recurring slots, and book the same 15 minutes on your calendar for next month. For the timing side of the audit, the guide to finding your best publishing times shows how to check your hours against your own data.
Google review
Trustpilot review
Product Hunt comment
