A social media audit does not require an agency, a tool subscription, or a week. It requires one honest hour, your own profiles, whatever native analytics each platform provides, and a blank spreadsheet. At the end you will have a one-page scorecard and, more importantly, one decision per channel: kill it, keep it, or double down.
The audit runs as five timed passes. The timings below assume three to five active channels; with more, borrow time from the profile pass, not from the content mix pass, because the content mix is where the real findings live.
Set up the scorecard (5 minutes)
- Create a spreadsheet with one row per channel you have ever posted on, including the abandoned ones.
- Add columns: profile score, positioning score, content mix notes, engagement read, decision.
- Score crudely on 0 to 2 throughout: 0 missing, 1 present but stale, 2 current. Precision is not the goal; comparability is.
- Open each platform's native analytics in a tab now so the later passes do not stall on logins.
Pass 1: profile completeness (10 minutes)
Visit each profile as a logged-out stranger would. You are checking whether someone who lands here for the first time can tell what you do, who it is for, and where to go next.
| Check | What a 2 looks like |
|---|---|
| Handle and display name | Consistent with your other channels; findable by searching the company name |
| Bio | Names who the product is for and what changes for them, in plain words |
| Link | Works, and lands on a page that matches the bio's promise |
| Pinned post | Current, and the post you would want a first-time visitor to read |
| Avatar | Legible at 32 pixels; the same mark across channels |
| Banner or header | Not announcing something that already happened |
| Contact route | An email, form, or DM setting through which a customer can actually reach you |
Pass 2: positioning consistency (10 minutes)
Write your one-sentence positioning from memory before looking at anything: who the product serves and what it changes for them. Then read each channel's bio and its last five posts and ask one question: could a stranger reconstruct that sentence from what they see here? Score each channel 0 to 2 and note the drift you find.
Drift is normal and usually invisible from the inside. The bio written at launch describes a product that has since moved on; the X profile leans on a joke that made sense in one conversation; the LinkedIn page still targets an audience you pivoted away from. The audit's job is only to see it. Fixing it is a ten-minute task per channel once seen.
Drift, side by side
LinkedIn bio: 'AI content workflows for B2B teams.' X bio: 'building fun marketing things.' Website: 'the social scheduler for solo founders.' Three claims, no overlap.
Consistent, adapted
LinkedIn: 'Helps founders turn product work into a week of social posts.' X: 'turning changelogs into posts people actually read. built in the open.' Same claim, channel-native delivery.
Pass 3: the last-30-posts content mix (20 minutes)
This is the longest pass and the one that pays for the hour. On your primary channel, tag each of the last 30 posts with the job it was doing: teach, perspective, proof, progress, promo, or other. If you post less often, tag the last 30 days across all channels instead. Count the tags and compare against the target ranges below, which come from the four-pillar founder calendar.
| Post job | Target share | Too little looks like | Too much looks like |
|---|---|---|---|
| Teach | 30 to 40% | Followers get nothing they can use without buying | A free course with no reason to remember who wrote it |
| Perspective | 15 to 25% | A feed with no point of view to agree or argue with | Opinions with no product or proof underneath them |
| Proof | 10 to 20% | Claims that never show evidence | A highlight reel that reads as bragging |
| Progress | 15 to 25% | The audience forgets you are building something | A changelog nobody outside the team can follow |
| Straight promo | 0 to 10% | Nobody ever learns how to buy | The feed reads as ads and the reply count shows it |
Worked example (hypothetical)
A founder tags 30 posts and counts: 3 teach, 2 perspective, 0 proof, 17 progress, 6 promo, 2 other. The diagnosis reads itself: the feed narrates the roadmap and asks for signups, but gives a stranger no reason to follow. The next month's fix is written in the gaps: one teaching post per week, and the first proof post as soon as a customer result can be shared honestly.
Pass 4: engagement quality (10 minutes)
Ignore raw like counts; they are the easiest number to inflate and the hardest to act on. Over the same 30 posts, count the things that indicate a human decided you were worth their time.
- Replies from real accounts that are not teammates or friends.
- Questions asked to you, in comments or DMs; save the exact wording, it is next month's content.
- Mentions or shares that reference what you said, not just your handle.
- Link clicks and profile visits where native analytics show them.
Pass 5: kill, keep, or double down (5 minutes)
Classify each channel from pass 4 as conversation (real replies recur), broadcast (clicks or reach without discussion), or silence (neither). Then take the decision. This is the pass founders skip, which is why dead channels stay on life support for a year.
| What the scorecard shows | Decision | What it means in practice |
|---|---|---|
| Silence, and no strategic reason to be there | Kill | Park the handle, point the bio at the channel that works, stop scheduling into a void |
| Silence, but your buyers demonstrably live there | Keep, on probation | Change the content mix before you blame the channel; re-audit in 60 days |
| Broadcast with steady clicks | Keep | Serve it with adapted content at low cost; stop chasing conversation that is not coming |
| Conversation, even at modest numbers | Double down | Replies are the scarce signal; give this channel your original material and your reply time |
From audit to next month
An audit that does not change the next 30 days was an hour of self-inspection. Carry exactly three fixes into next month's calendar: the biggest profile gap, the most lopsided content ratio, and the channel decision you have been avoiding. Then put a quarterly repeat in the calendar; the second run takes less than an hour because the scorecard already exists.