Free Twitter Account Audit

How to audit a Twitter account

An audit of a company's Twitter account answers three questions from outside: is the account posting regularly, are the posts landing with the people who see them, and does the content say the same thing every time? You do not need access to the account. Everything this tool reads is visible under each post on X: its views, likes, replies, reposts and quotes.

1. Collect the last ten original posts. Skip replies and reposts: their reach is tiny and drags every number down. Skip posts from the last two days, which have not collected their engagement yet.
2. Enter them, paste them from a spreadsheet, or import a CSV. Add the post text if you want the content checks, and a few replies if you want the reply checks.
3. Analyze. You get the posting rhythm, the engagement rate per post, and a card for every signal the sample shows, each with its evidence, the line it was judged by, and what it cannot tell you.

Twitter marketing signals to look for

Inconsistent posting

Long silences between bursts of posts. The audit compares the longest gap in the sample with the account's own median interval: an 18-day gap means something for an account that usually posts every three days, and nothing for one that posts monthly. It also flags an account that has not posted for a while, measured against the same rhythm. A gap is an observation, not proof that anyone stopped caring: holidays, launches and team changes all cause them.

Low engagement

Posts that are seen but not answered. The audit measures engagement against views, post by post, and only calls it a signal when most posts fall below the line, not one quiet post. When a typical post gets over a thousand views and still almost no replies, the account has an audience and is not converting its attention: the clearest opening for an agency.

Repetitive content

The same message posted again and again: near-duplicate posts, the same call to action in most posts, the same hashtags on every post. The audit compares the words of every pair of posts and counts repeated phrases. Consistent branding is not the problem; a feed where every post asks the reader to do the same thing is.

Negative reply patterns

Replies that complain: support questions, refunds, things not working. The audit only looks at replies you paste in, and only reports that complaint-type language appears often enough to review by hand. It cannot see whether the company answered in private, and it never draws conclusions about the company from a few words.

How to analyze Twitter engagement

The engagement rate by views is likes plus replies plus reposts plus quotes, divided by views, times 100. It is calculated for each post that has a view count. The audit reports the median as well as the average, because one viral post can drag an average far from what a typical post gets; the median is what the account can expect.

engagement rate = (likes + replies + reposts + quotes) / views x 100

Per view, not per follower: on X most of what a post reaches is decided by the feed, not by who follows the account, so dividing by followers mostly measures the algorithm. Enter the follower count above and the audit shows the per-follower rate too, as a second figure: interactions per post divided by followers. X Analytics shows the account owner a different rate that also counts clicks and profile visits, which outsiders cannot see; this audit's number is lower and works on any account. There is no universal "good" rate: compare an account with its own past and with accounts like it, and use the thresholds above to set your own lines.

How marketing agencies can use Twitter audits

• Qualify before you write. A company with a long posting gap or high views and no replies has a visible problem an agency can fix. A healthy feed is a weaker lead.
• Open with the evidence. "Your last ten posts average about 3,000 views and two replies" is a first line a founder recognises. A generic audit offer is not.
• Re-run it later. The same sample a month after you start is the simplest before and after for a client.

Coldpitch runs this kind of reading on every company it finds, alongside signals like funding, hiring and launches, and drafts the first message from what it saw.

Limitations of public Twitter analytics

• A view is not a person. X counts impressions, so one person can be several views.
• Visible engagement is not all engagement. Clicks, profile visits, bookmarks and shares by message are invisible from outside.
• Ten posts are a sample. They show the last few weeks, not the account's history. Confidence on each card says how complete the data was.
• Causes are not visible. A spike or a drop is an observation; the audit lists possible reasons but cannot tell which one it was, and it never concludes that engagement was bought.