A social media content audit takes 20 minutes if you do one thing: compare every post to your own median, then sort the results into keep, fix or retire. You do not need to review every post you ever made. Thirty posts and a clock are enough.
TLDR: Pull the last 30 posts, find your median, then group posts by format and topic and mark each group keep, fix or retire. Posts below 0.8x your median are your low performers. In Mydrop's data, carousels ran 10% above an account's typical post, which is a good first thing to look for.
Minute by minute
| Minutes | Step | What you produce |
|---|---|---|
| 0 to 3 | Pull the last 30 posts with reach, engagement, format, date and hour | One table |
| 3 to 6 | Find your median reach and engagement rate | Your baseline |
| 6 to 10 | Mark each post as above 1.1x, near 1.0x or below 0.8x | A color-coded list |
| 10 to 15 | Group by format and by topic, and note which groups cluster high or low | Two or three patterns |
| 15 to 18 | Label each group keep, fix or retire | A decision per group |
| 18 to 20 | Write the next month's changes | A short action list |
Step 1: use your own median as the yardstick
An audit against someone else's benchmark tells you little. Use your account's median: sort your 30 posts by reach and take the middle value. A post at 0.9x the median is noise. A post at 0.6x is a real miss.
For a sense of normal range, in Mydrop's September 2026 data the median Instagram post earned a 3.1% engagement rate, and the middle half of posts ranged from 1.1% to 6.7%. If your numbers sit inside that range, the account is typical.
Step 2: look for patterns, not single posts
One bad post is bad luck. Three bad posts that share something is a finding. Check these in order:
- Format. Which formats sit below 0.8x? In our data, carousels ran 10% above an account's typical post, photos and videos sat at the median and a link in the caption cost 7%.
- Time. Do low posts cluster late at night? Posts published between midnight and 6 am in local time ran 17% below the median across platforms.
- Topic. Does one topic keep landing low even in different formats?
- Caption. In the same data, captions under 80 characters ran 9% below median. Longer ones were within 3% of it, so length was not a lever past that point.
Step 3: sort each group into keep, fix or retire
Here is the rule for low performing content:
| Result | What it means | Action |
|---|---|---|
| Group above 1.1x median | Working | Keep and add one slot a month |
| Group between 0.8x and 1.1x | Average | Fix: change one thing (format, hour, hook) and re-test |
| Group below 0.8x after 3 or more posts | Not working | Retire it, or change it once and re-check |
Change one thing at a time when you fix, or you will not know what worked.
Step 4: write down the changes
End every audit with a short list of changes for the next four weeks. Keep it to three:
- One group to expand.
- One group to change.
- One group to stop.
Put the changes into your calendar so they happen by default and not by memory. In Mydrop's calendar you can set them as recurring slots per account.
What to skip
Not everything needs auditing in 20 minutes. Leave these for a longer review:
- Old posts from before your strategy changed.
- Paid posts, which have their own metrics.
- Comparison with competitors, which takes more time and often shows less than your own data.
Make the audit faster each month
Most of the 20 minutes goes on pulling and sorting numbers. Mydrop's performance view shows each post against that account's own median, so steps 1 and 2 shrink to a few minutes, and scheduled reports can send the summary to you each month. If you manage several accounts, that is the difference between an afternoon and a coffee break.
For a content audit built around AI answers and search, see the AI-ready content audit. If your audit shows reach is stuck, our 30-day recovery plan turns the findings into weekly steps.
Next step: set a 20-minute timer, pull your last 30 posts and find your median. You can start free with Mydrop and connect your accounts to see every post against its own baseline.
Google review
Trustpilot review
Product Hunt comment
