Monthly agency reports have a characteristic shape: many pages, several charts, a lot of numbers that go up, and no clear answer to whether the money produced anything.
This is not always deliberate. Reporting on activity is easier than reporting on outcomes, particularly where outcomes are slow. But it means the client cannot judge the relationship, which over time suits one party more than the other.
Here is what a useful report contains.
What was actually done
A plain list of the work completed during the month. Not "ongoing optimisation" -- specific items. These four pages were written. This listing was corrected. These images were compressed. This campaign ran with this budget.
You should be able to go and look at most of it. If nothing on the list results in something you can inspect, ask what you are paying for.
Enquiries, by source
The number that matters. How many enquiries arrived, and where from.
Everything else on the report is a means to this. An agency that reports impressions and traffic but cannot tell you about enquiries is measuring its own activity rather than your outcome.
If attribution is genuinely difficult in your case, that should be discussed honestly and a method agreed -- tagged links, distinct numbers, an intake question. Difficulty is a reason to solve it, not to omit it.
The comparison
Every figure with a like-for-like comparison: previous month, and the same month last year if the business is seasonal.
An isolated number tells you nothing. The direction is the information.
What did not work
The section that distinguishes a good agency from a competent one.
Something every month underperforms. A report where everything succeeded is not describing reality, it is managing you. Expect a paragraph on what was tried, did not work, and what will be done differently.
An agency willing to write that is one you can trust with the rest.
Next month, specifically
What will be done, in the same concrete terms. This lets you check next month whether it happened, which is the only way a monthly relationship stays honest.
The numbers that are there to fill space
Impressions. How many times something was theoretically displayed. Enormous numbers, minimal meaning.
Reach and followers. Only relevant if they lead to enquiries, which should be measured directly.
Bounce rate, presented without context. Frequently misinterpreted, and a high figure on a page that answered the question is not a problem.
Keyword rankings in isolation. Ranking well for phrases nobody searches, or that do not lead to purchase, is not an achievement. Rankings are worth reporting alongside whether traffic and enquiries followed.
Anything expressed only as a percentage increase. A three hundred percent rise from one to four is technically accurate and misleading. Insist on absolute numbers.
The meeting to have quarterly
Monthly reports track execution. Every quarter, a different conversation: is this working, in business terms.
Enquiries and customers attributable to the work, set against the cost. Whether the trend justifies continuing. What would need to change to justify spending more, or less.
Agree at the start of the relationship what success looks like and when it will be judged. Relationships without a review point tend to continue indefinitely regardless of results, because nobody has a natural moment to raise it.
A reasonable expectation of pace
Some of this work is slow, and a report showing little movement in month two is not evidence of failure.
What should be visible from the beginning is the work itself. Pages written, listings fixed, technical faults corrected. If months pass with no artefacts and no enquiries, the question is not whether to be patient. It is what is actually happening.
The report you can write yourself
If your agency will not provide this, keep your own version: enquiries received each month, where they came from, how many became customers, and what you paid.
Four numbers, updated monthly. That is enough to know whether any marketing spend is worth continuing, regardless of what anyone else reports.