What your monthly marketing report should say

Most monthly reports we see when a client moves to us are activity logs. Twelve pages of what the agency did, several charts with no comparison, and a summary paragraph that could have been written before the month started.

They're not dishonest. They're just answering the wrong question. You don't need to know how busy someone was. You need to know whether the money worked.

Here's what a report should contain, and how to tell quickly if yours doesn't.

1. The agreed number, against its baseline

One metric, named at the start of the engagement, shown next to where it was before.

Not "organic traffic is up". Up from what, over what period, compared with the same period last year? A 20% rise in December means nothing until you know December is always up 25%.

If the report opens with impressions, reach or "visibility", that's a choice. Those numbers go up almost regardless of what anyone does, which is precisely why they lead so many reports.

2. What changed, and why someone thinks that

The hardest part and the part most often missing. Not a list of what moved — an actual explanation.

Enquiries fell 18% this month. Two of our three best-performing pages dropped positions after the September core update, and the traffic they lost converted well. We've prioritised those two for the next cycle.

That paragraph is worth the whole rest of the document. It commits to a cause, which means it can be wrong — and an agency willing to be wrong in writing is an agency actually thinking.

Compare it with "performance fluctuated due to algorithm changes". That sentence has been in every report since 2011 and means nothing.

3. What didn't work

A month with no failures is a month nobody tested anything.

A useful report names the things that were tried and didn't pay off: the ad variant that lost, the page that didn't rank, the sequence with a worse reply rate. Partly because it's honest, mostly because it's the only evidence that anyone is running experiments rather than maintaining a routine.

If you have never once read a negative result in a report you've paid for, that's worth asking about directly.

4. What happens next, specifically

Named actions, with owners, before the next report.

"Continue optimising campaigns" isn't a plan. "Rewrite the three service pages that lost positions, publish by the 14th" is. The second one can be checked next month. The first exists so that it can't be.

Watch for whether last month's "next steps" ever get mentioned again. In a lot of reports they quietly vanish, which tells you they were written to fill a slide.

5. What the agency needs from you

Almost always absent, and almost always relevant.

Engagements stall on the client side more than anyone admits — approvals that take three weeks, an expert who's never free, access that never arrives. A report that says "we're blocked on sign-off for the two landing pages, sent 3 October" turns a delay into a shared problem rather than a surprise three months later.

Four questions that expose a weak report

Ask these at the next review:

1. "What would have to be true for us to stop doing this?" — tests whether anyone has defined failure.
2. "Which of last month's actions didn't work?" — tests whether experiments are real.
3. "How does this compare to the same month last year?" — tests whether seasonality is being handled or quietly used.
4. "If our budget halved, what would you cut first?" — the most revealing of the four. An agency that knows which half of the spend is weakest has been paying attention. One that says "it all works together" has not.

The honest caveat about length

A short report isn't a lazy one. We'd rather send one page that answers the four questions above than twelve that answer none of them — and a long report is sometimes a deliberate way to make a flat month feel busy.

Judge it by whether you could make a decision from it. If you read your report and still don't know whether to spend more, spend less, or change direction, it hasn't done its job regardless of how many charts it contains.

Getting the measurement right so the report can be honest is where most of our analytics work starts, usually before anyone writes a report at all.