Most ecommerce founders treat their LinkedIn like a treadmill. Post, post, post, never look back. They'll write 20 posts a month and never once ask which 4 did all the work.
That's the gap. We run a monthly content retro for every client, and it's the single highest-leverage 30 minutes in the whole system. Not because it produces new content — it doesn't. Because it tells you what to stop making.
Here's the system, exactly as we run it.
Why posting volume without review is a trap
Volume feels like progress. It isn't. We've audited founder accounts pushing 25 posts a month where 18 of them got under 800 impressions and the founder had no idea, because the dashboard only shows you the last post you opened.
When you don't review, three things happen. You keep repeating formats that don't land because they're comfortable. You abandon formats that worked because you forgot they worked. And you burn your best ideas on Tuesdays at 4pm when half your audience is offline.
A retro fixes all three. It turns a pile of disconnected posts into a signal. The point isn't to grade yourself. The point is to find the 20% you should kill and the 20% you should double.
The 30-minute monthly retro, step by step
Block 30 minutes on the first Sunday of the month. Same time every month — make it a recurring calendar hold, not a "when I get to it." Here's the run of show.
Minutes 0–10: Pull the raw numbers. Open LinkedIn analytics, export the last 30 days of posts into a simple sheet. Five columns only: post hook (first line), format (text / carousel / image / poll), impressions, engagement rate, and DMs or profile actions you can trace to it. That last column is the one that matters. Impressions are vanity. Conversations are pipeline.
Minutes 10–18: Rank and band. Sort by your real success metric — for most founders that's a blend of engagement rate and inbound conversations, not raw reach. Then band the posts into three groups:
- Top 20% — the ones that drove comments from your actual buyers or landed in your DMs.
- Middle 60% — fine, forgettable, did their job.
- Bottom 20% — under-performed your median by a wide margin.
Minutes 18–26: Find the pattern in each band. This is the actual work. Look at your top 20% and ask: what do they share? Same hook type? Same format? Same topic pillar? Same posting day? We almost always find that a founder's winners cluster around one or two angles they didn't realize were their strongest.
Then do the same for the bottom 20%. Usually it's one of four things: a format that doesn't fit your audience, a topic too far from your offer, a hook that buried the point, or a vague "thought leadership" post with no specific number in it.
Minutes 26–30: Write two decisions. Not ten. Two. One thing you'll do more of next month, one thing you'll stop. Write them at the top of next month's content doc so they're staring at you when you plan.
That's the whole system. Thirty minutes. The discipline is in the cadence, not the complexity.
The kill list: what your bottom 20% is telling you
Founders hate killing posts. They get attached to a format because they spent two hours on a carousel that flopped. The retro removes the emotion — the data makes the call for you.
Across the accounts we manage, the bottom 20% almost always falls into these buckets:
- The format mismatch. Polls that get votes but no DMs. Carousels nobody saves. If a format produces engagement but never a conversation, it's entertainment, not pipeline.
- The off-target topic. A post about your weekend, a generic motivation take, a hot take on a topic your buyers don't care about. Engaging to strangers, invisible to prospects.
- The buried hook. Good idea, terrible first line. The retro catches these because you can see a strong topic that died on a weak open — that's a fixable, not a kill.
- The no-number post. Vague claims with nothing to anchor them. Our data is consistent: posts with a specific number in the first two lines outperform vague ones by a wide margin on both engagement and saves.
You're not deleting these posts. You're refusing to make more of them.
The double-down list: scaling your winners without repeating yourself
The mistake founders make with winners is they copy them flat. Same post, slightly reworded, three weeks later. Audiences notice and it decays fast.
Doubling down means scaling the pattern, not the post. If your winner was a contrarian take with a client number, you don't rewrite that exact take — you find three more contrarian takes you can back with numbers. If your winner was a "here's what I'd do differently" teardown, you make that a recurring format with fresh subjects.
We tell clients: a winner is a template, not a script. Pull out the structural reason it worked and apply it to new material. That's how one good post becomes a content pillar instead of a one-off.
The other half of doubling down is timing. If your top three posts all went out Tuesday and Thursday mornings, that's not a coincidence — that's your audience telling you when they're around. Move your best material into those windows on purpose.
FAQ
How many posts do I need before a retro is useful? At least 12 in the month. Below that, the bands are too small to see a pattern. If you're posting twice a week, run the retro every two months instead.
What if I don't have DM or pipeline data? Use the closest proxy: meaningful comments from people who look like your buyers, profile views in the days after a post, and saves. Saves are an underrated signal — they mean someone wants to come back to it.
Should I do this weekly instead? No. Weekly is too noisy — one viral post or one dead week skews everything. Monthly gives you enough volume to see real patterns without overreacting to randomness.
Does this replace planning my content? It feeds it. The retro's two decisions become the inputs for the next month's plan. Retro first, then plan. Never plan in a vacuum.
The point of looking back
The founders who win on LinkedIn aren't the ones posting the most. They're the ones who compound — every month their content gets a little sharper because they actually looked at what happened.
Thirty minutes a month. Kill the bottom 20%. Double the top 20%. Do that for six months and your account stops being a treadmill and starts being a system.
If you'd rather have someone run the retro, find the patterns, and feed them straight back into your content, that's a core part of what we do at EcomGhosts. We don't just write — we measure, prune, and reinvest, so your authority compounds instead of plateaus.