A social content flywheel is a weekly loop in which each post teaches you what to make next: publish, measure against your own median, pick the top performers, remake them in a new format and publish again. It does not rely on trends because the ideas come from what your own audience already responded to. Trends become a bonus, not the engine.
TLDR: Every week, score your posts against your median, take the top 10%, remake each one in a new format or angle, and schedule it again. In Mydrop's data the gap between the best and worst post format was only 11%, so the loop matters more than the format.
The loop in five steps
| Step | What you do | Time per week |
|---|---|---|
| 1. Publish | Post your normal schedule, tag each post's topic and format | Your usual |
| 2. Measure | Score each post as engagement rate divided by your median | 15 minutes |
| 3. Pick | Take the top 10% of posts from the last 30 days | 5 minutes |
| 4. Remake | Turn each winner into a new format or a new angle | 1 to 2 hours |
| 5. Reschedule | Publish the remake 3 to 6 weeks later, to a mostly new audience | 10 minutes |
The speed comes from step 4. You are not starting from a blank page, because a post that already worked is your brief.
Step 2: measure against your own median
A post at 4% means nothing without context. Divide it by your median. In Mydrop's September 2026 data, the median Instagram post earned 3.1% and the middle half ran between 1.1% and 6.7%. If your median is 2%, a 3% post is a 1.5, a clear winner.
Use at least 30 posts for the median. With fewer, one lucky post moves it too much.
Step 3 and 4: remake the winners, not copy them
Posting the same thing again feels lazy and sometimes it is. The aim is same idea, new packaging. Pick one of these for each winner:
- Change the format. A strong text post becomes a carousel. A carousel becomes a short video. In our data the spread between formats was only 11%, with carousels highest at an index of 1.10 and videos at 0.99, so the format change adds a different audience, not a guaranteed lift.
- Change the angle. A "5 mistakes" winner becomes "the one mistake that costs the most".
- Go one level deeper. Take the question people asked in the comments and answer it in a full post.
- Change the platform. A LinkedIn winner can be reshaped for Instagram, but check each platform's limits first.
Step 5: wait before you post it again
Do not repost within a few days. Most of your followers will remember it. A gap of a few weeks lets the same idea reach people who missed it. Use Mydrop's calendar to schedule remakes ahead of time and spot weeks where the plan is empty.
Add a second loop for evergreen posts
Some posts keep earning. Mark those as evergreen, and put them into a library that you re-run every quarter. This is the slow wheel next to the weekly one. It protects you in weeks when you have no time to make anything new.
For a fuller plan, see how to stop chasing virality and focus on evergreen growth.
Where trends fit
Trends still work. Treat them as a side bet: about one post in five on a trend that fits your brand, and four from the flywheel. The flywheel gives you a stable base. If a trend post wins, it goes into the loop like any other winner.
What breaks a flywheel
- No median. Without a baseline, you cannot tell winners from luck.
- Skipping step 4. You keep measuring and never remake anything.
- Too many variables. Remake one thing at a time, so you know what worked.
- Making it a one-person habit. Put the weekly 30 minutes in the calendar for the whole team.
Mydrop's performance view does step 2 for you: it lines each post up against the account's own median across profiles, so picking the top 10% takes minutes instead of an afternoon in a spreadsheet.
Remake your top post of the month in a new format
Pull the last 30 days of posts, score each against your median and mark the top three. Remake one of them in a new format this week and schedule it for three weeks from now. To get the scoring built in, start free with Mydrop.
Google review
Trustpilot review
Product Hunt comment
