Most social media audits die of scope. Somebody downloads a template with forty tabs, blocks out a full day, gets through two platforms, and the spreadsheet joins the graveyard of half-finished strategy documents. Six months later the same problems are still there, unexamined.
The fix is a timebox. Two hours, a timer, and the discipline to skip everything that does not lead to a decision. You will not learn everything about your accounts in two hours. You will learn the ten things that matter, which is more than most teams know right now.
This is the pass I run on my own accounts every few months. It has five steps, each with a time budget, and the budget is part of the method. When the timer goes off, you move on.
Before the timer starts
You need three things ready, or the first half hour disappears into logins.
Access to native analytics on every platform you post to. Meta Business Suite, LinkedIn analytics, X analytics, TikTok Studio, YouTube Studio, whichever apply. Test the logins the day before, not during the audit.
A blank spreadsheet. Not a template. Templates tell you what someone else thought was worth measuring, and half the columns will not apply to you. You will build maybe six columns as you go.
A fixed window. I use the last 90 days. Long enough to smooth out one lucky post, short enough that the platform algorithms and your own strategy have not shifted underneath the data.
Step 1: inventory every account (20 minutes)
List every social account your business owns. All of them, including the ones you are embarrassed about. The Pinterest from 2021. The TikTok someone started at an offsite. The second Instagram created for a campaign that ended.
For each one, record four things: platform, handle, last post date, and who has the login. That final column is the one that produces surprises. Accounts where the only person with access left the company are a real and common finding, and you want to discover this on a calm Tuesday rather than during an incident.
Anything that has not posted in 90 days goes on a watchlist. You are not deciding its fate yet, that comes in step 5. Right now you are just making the map honest.
Step 2: the profile pass (20 minutes)
Open each active profile the way a stranger would, logged out or in an incognito window, and look at what they see. Not what you meant. What is actually there.
Check four things on each profile:
- Does the bio say what you do in plain words, or is it a slogan that only makes sense to people who already know you?
- Does the link work, and does it point where you currently want traffic? Old campaign pages hide here for years.
- Is the pinned post recent and representative, or is it the launch announcement from a different era of the company?
- Do the name, handle, and avatar match across platforms closely enough that someone could confirm it is you?
Fix the trivial stuff immediately. A dead link takes ninety seconds to repair and there is no reason it should wait for a report. Everything else goes in the spreadsheet.
Step 3: read 90 days of content (40 minutes)
This is the longest block and the one where discipline matters, because analytics dashboards are built to be wandered around in. You are not wandering. You are answering three questions per platform.
First: what were the top five and bottom five posts by reach? Skim them side by side and write one sentence about the pattern you see. Format, topic, tone, day, whatever jumps out. Do not overthink it. Patterns that need a statistician to detect are not patterns you can act on with next week's content.
Second: re-sort by link clicks, or whatever action is closest to your business, and look again. This ordering is usually different from the reach ordering, and the difference is the most useful thing in the whole audit. The post that reached the most people and the post that made people act are rarely the same post. If you want to go deeper on that distinction, I wrote a full piece on which social metrics actually matter.
Third: what does the posting cadence actually look like? Not what the strategy doc says. Count the real gaps. Three posts in one week and then silence for two is a pattern worth knowing about, because consistency problems usually masquerade as content problems. If cadence turns out to be your finding, how often to post on each platform covers what a sustainable floor looks like.
Ten minutes per platform is enough for this if you resist the urge to build charts. Sentences in a spreadsheet, not charts.
Step 4: follow your links (20 minutes)
Now click every link you currently publish. The bio links, the links in recent posts, the link in the newsletter signup CTA. All of them, as a user.
You are checking two layers. The first is basic function: does the link resolve, is the page it lands on current, does it make sense on a phone. Broken and outdated destinations are embarrassingly common because nobody clicks their own links after launch day.
The second layer is measurement. When someone clicks a link from a post, can you tell? Later, can you tell what they did after clicking? If every post points at a bare URL with no tracking, the honest answer is no, and your step 3 analysis was running on reach because reach was all you could see. Adding UTM parameters to your social links is the cheapest fix, and it changes what your next audit can tell you. This is also the layer where a scheduler that creates tracked links per post earns its keep, though the audit itself needs no tools beyond a browser.
Write down, per platform: link works or broken, destination current or stale, click trackable or invisible.
Step 5: write the decision list (20 minutes)
An audit that produces a report produces nothing. The output of these two hours is a short list of decisions, and I mean short. Seven or fewer, written as actions with owners and dates, sorted into three buckets.
Stop. The abandoned accounts from step 1 get an explicit fate: delete, or post a pinned "find us here" and let them rest. The format that has been at the bottom of the performance list for 90 days stops. Killing things is the highest-return part of the audit and the part people skip.
Keep. Name what is working so it does not get accidentally dropped in the next redesign of the content plan. The format that keeps ending up in the top five is now officially a series, with a slot on the calendar.
Try. One or two experiments, not five. Each one framed so that the next audit can judge it: what you are trying, on which platform, and what number should move.
Then, and this is the part that separates audits that change things from audits that get filed: put the decisions on the actual content calendar. A decision without a scheduled slot is a wish.
What to deliberately leave out
Two hours works because of what it excludes. Competitor teardowns are a separate session, and a useful one, but mixing them into an audit turns observation into anxiety. Follower demographics are interesting and rarely actionable for a small team. Sentiment analysis tools, engagement-rate benchmarking against industry averages, brand share of voice: all of it can wait until the basics above are handled, and for most teams the basics are not handled.
The uncomfortable truth I keep relearning on my own accounts is that step 4 finds the worst problems. Content quality gets all the attention, but the dead bio link and the untracked clicks are usually costing more than the mediocre caption ever did.
Run the audit, fix the links, schedule the decisions. Then put a recurring reminder in your calendar for 90 days out, because the second audit is where this gets genuinely useful: it is the one with a baseline.