HASHTRACKER

X and Twitter analytics 2 min read

Which Twitter stats are worth tracking every month

Most Twitter stats rise simply because you posted more. Impressions per post is the line that says whether an account is improving, and almost nobody plots it.

Cover drawn as a column chart

The analytics tab shows plenty of Twitter stats and answers very little. Here is what to pull out of it, and what it cannot give you at all.

The four figures that describe an account

Posts published, impressions earned, engagement collected and impressions per post. Over a fixed period, with the previous period next to them. The fourth is the one that matters. Total impressions rise simply by posting more, which tells you nothing about whether the account is getting better.

We can pull this for any account, including ones that are not yours, within the Twitter compliance rules.

We can pull this for any account

Why posting more is not improvement

An account that doubles its output and doubles its impressions has stayed exactly where it was. Each post is doing the same work. Track impressions per post over months and the trend becomes visible. Rising means the account is being served to more people per post. Falling while volume climbs usually means the audience is tuning out.

That single line is worth more than any monthly total.

Separating your reach from other people's

Split impressions into your own posts and posts by others that mention you. A brand where nearly all the reach comes from its own account is broadcasting. One where a large share comes from other accounts has a conversation happening around it.

Both are legitimate positions. Not knowing which one you are in is not.

What your best posts have in common

Sort by impressions, take the top twenty and look for the pattern. Format, length, time of day, whether there was a link, whether it was a reply. The answer is often boring and actionable: replies outperform originals, or posts without links travel further.

Do the same for the bottom twenty. The contrast is where the instruction is.

Account age and posting history of the people around you

For an account you are studying rather than running, the author profile of everyone engaging with it is revealing. If most of the accounts amplifying a brand were created in the last six months and post constantly, that is a different audience from one made of long standing accounts that post occasionally.

Nobody needs a bot label to draw a conclusion from that. The account creation dates and post counts are enough.

Getting the figures for an account you do not own

Your own analytics cover your account only. Everything above can be produced for any public account through the API, with the same impression figures the platform reports. Pick a plan and the report arrives with the dataset, so every figure traces back to the posts it came from.

Pick a plan

Choosing a subscription

Every subscription carries the same product. The only decision is how many tweets you need to analyse.

See the subscriptions