The Four Numbers Worth Reporting Monthly

There is a particular kind of monthly report that runs to twenty-two pages, opens with a bounce rate chart, includes a word cloud, and gets filed unread. It is long because length reads as diligence. It is unread because none of it connects to a decision anyone is about to make.

A better test for any figure in a report: if this moved, what would you do differently? Most metrics fail that test. Four survive it, and they happen to form a chain, which is what makes them useful together rather than merely individually.

Why the long version fails

Two reasons, and they compound.

The first is that a dashboard with forty tiles has no argument in it. Everything is present, so nothing is emphasised. The reader has to work out what matters, which is precisely the job they were paying someone else to do.

The second is that a long report is easy to hide inside. When ten things are up and ten are down, any story can be told. Narrow it to four and there is nowhere to go — either they moved or they did not, and you have to explain why.

The four below run from earliest signal to the only one that pays wages. Each answers a different question, and the order tells you where a problem lives.

One: impressions

The question it answers: is your site being surfaced more often than it was?

Impressions count how many times your pages appeared in search results. Pull this from Search Console’s performance report on a rolling three-month view rather than calendar months, which smooths out the fact that February is short and August is quiet.

It comes first because it moves first. Fix a technical problem, publish a genuinely useful page, earn a link, and this responds well before traffic does. It is the earliest honest evidence that something is working — and equally, six months of flat impressions is the earliest honest evidence that something is not.

The caveat: rising impressions on their own are not success. They can rise because you started appearing in position forty for a hundred irrelevant phrases. That is why it is one of four rather than a headline.

Two: how many different searches you appear for

The question it answers: is your visibility widening, or just deepening?

Take the query list from the same report and count the distinct rows. Most reporting ignores this number entirely, which is a shame, because it distinguishes between two very different situations that look identical in a traffic chart.

A site appearing for two hundred phrases this quarter and three hundred and forty next quarter is being understood as relevant to a broader set of needs. That footprint is what durable growth looks like from the inside. A site whose count stays flat while impressions rise is simply climbing on the same small set of terms — worth having, but capped.

Track it as a plain count month on month. The direction is what matters, not the absolute figure, because sampling thresholds mean the number is approximate anyway.

Three: clicks arriving on pages that sell

The question it answers: is the traffic going anywhere that could produce revenue?

Not total sessions. Total sessions is the metric most likely to flatter and least likely to inform, because a popular blog post about a general topic can lift it substantially while the service pages sit exactly where they were.

Segment instead. Report clicks landing on the pages that exist to win work — your service pages, your location pages, your contact page — and hold the rest separately. If those page-level clicks are climbing, the site is attracting people with a commercial reason to be there. If the total is climbing while those pages are flat, you have written something popular rather than something profitable. Both are fine outcomes as long as you know which one you got.

This is also the number that exposes a title and description problem. High impressions on a service page with almost no clicks means people saw you and chose someone else in the results, which is a copy issue, not a ranking issue.

Four: enquiries

The question it answers: did the phone ring and the inbox fill?

Calls and form submissions, counted honestly. This is the only number in the report that has any relationship with the bank account, and it should be the largest thing on the page.

Two disciplines make it trustworthy. First, reconcile it against reality — the inbox, the call log, the diary — rather than trusting a dashboard figure that could be double-counting or dropping a share to consent tooling. Second, split it by rough quality if you possibly can. Thirty enquiries of which four are relevant is a worse month than twelve of which nine are, and a report that shows only the headline will tell you the opposite of the truth.

Where the business can attach a value to a closed job, do that too. Anything upstream of revenue is a proxy. This is as close to the real thing as a marketing report gets.

Reading the four as a chain

The order is the diagnostic. Work down it until something breaks.

Impressions flat? The site is not being surfaced. That is a visibility problem — technical, content or authority — and nothing downstream will move until it is addressed.

Impressions up, query count flat, clicks flat? You are being shown more but not chosen. Look at what the results page shows for your key terms, and at whether your titles say anything a person would want.

Clicks up, enquiries flat? The problem has moved onto the site itself. The wrong people are arriving, or the right people are arriving at a page that does not make it easy to get in touch. That is a website and messaging question, not a search one.

All four up? Keep doing what you were doing and resist the urge to redesign anything.

What belongs in an appendix

Nothing here is worthless. It is a question of billing.

Bounce rate and time on page describe behaviour without explaining it and are easily distorted by how tracking is configured. Average position moves constantly, differs by device and location, and tells you nothing about value. Social followers, impressions on posts and email open rates are channel housekeeping. Rankings for the handful of phrases the business owner personally types are the single most misleading number in the industry.

Put all of it in an appendix for anyone who wants it. Keep the front page to four figures, a sentence on what changed, and a sentence on what happens next.

If your current reporting runs long and says little, the fix is usually measurement that someone has actually checked, not another chart. Get in touch if you would like a second opinion on what you are being sent.