A month is not a unit of anything in marketing. It’s a unit of invoicing.
Almost everything that matters moves either slower than that or faster. Your organic traffic, your content, your position in a category: none of those change meaningfully in thirty days. Your spend, a broken form, a campaign running in the wrong geography: those change in an afternoon.
So a monthly document reports noise as though it were signal for the slow things, and arrives roughly three weeks late for the fast ones. It is the single cadence that is wrong for both ends of the business.
It survives because the invoice is monthly.
That’s the argument. And the useful consequence is not “write a better report”, it’s that you are asking one document to do two jobs it is badly shaped for, and a third job nobody admits it’s doing.
Two clocks, and one document trying to serve both
Split what you want to know by how fast it moves, and the whole thing falls apart neatly.
The fast clock is operational. Spend pacing, a form that stopped submitting, a disapproved ad, a campaign that quietly started serving nationally. What that needs is an alert on the day, from whoever is in the account. Nobody has ever been well served by learning on the 8th of November that their contact form broke on the 4th of October, and the chart showing the dip is not a report, it is an autopsy.
The slow clock is strategic. Whether the channel works, whether the positioning is landing, whether the pipeline is building. One month of movement in any of those is inside the noise. You need three or four months before the number means anything, which makes a quarterly judgment the right shape, and a monthly one actively misleading, because thirty days of random variation presented in a chart looks exactly like a trend.
Between those two there is very little left that a month genuinely measures.
What a month is actually good for
There is something, and it’s the section almost no report contains.
A month is a clean unit for reconciliation and for commitments. What we said we would do, what we actually did, what it cost. Three facts, one paragraph each, and you could write the whole thing on one screen.
Last month we said we would launch the two service pages, restructure the branded campaign and fix the tracking on the quote form. Two of those happened. The third did not, because your developer was unavailable, and it is now the first thing this month. Spend was $8,400 against a $9,000 budget. Here is the one number we watch, here is what it did, and here is whether that changes anything.
Then the only sentence that makes it a report rather than a record: what we are doing differently as a result, or explicitly that nothing changes and why.
If nothing in the document could have changed what happens next month, it is a timesheet with charts. That’s the test I would apply, and most reports fail it in the first three pages.
Which numbers belong in it is a separate argument that I have already had at length in why impressions are still in your report, and the questions to ask when the answers sound too smooth are in what to ask when an agency’s numbers go up. I won’t re-run either.
Your report is long, and it is not laziness
Here’s the part the client side gets wrong, and it’s worth being fair about.
You are paying for work you cannot see. Not out of distrust, just structurally: the work happens in accounts you don’t open, in documents you don’t read, in decisions made on a Tuesday you weren’t part of. The report is frequently the only artefact you ever receive.
So the document is carrying a job nobody wrote down, which is proving that work exists. That’s why it’s long. The length is a response to a real anxiety, and the person producing it is not padding out of vanity, they are answering a question you have never asked out loud.
Which is why a shorter report, introduced on its own, reads as less work. Any agency that has tried it has had the call. You cannot unilaterally shorten the document while it is still the only evidence in the relationship.
And this got worse this year, quietly. Reporting is the first thing agencies automate, and generating a long report now takes minutes rather than hours. The time-saving figures in circulation come from the companies selling the tools, so I would not quote one at you, but the direction is not in dispute. Length used to be weak evidence of effort because it cost somebody an afternoon. Now it costs nobody anything, so the signal is gone while the document keeps growing.
So the fix is not a shorter report
If length is doing the job of proving work exists, take that job away from it and give it to things that do it better.
Direct access to the accounts, in your own name. Not a dashboard somebody built for you, the actual Google Ads and GA4 accounts, with you holding Admin. That single change does more than any reporting format, because the question “is work happening” becomes answerable without asking anybody. I went through how to check what you actually own when writing about inherited accounts, and the ownership part applies while the relationship is healthy, not only when it ends.
A visible commitment list that both sides can see between meetings. Not a project plan. Five lines, updated, showing what is in flight.
And a standing call where the decision gets made out loud. The report should be the artefact of that conversation rather than a substitute for it.
Do those three and the document can shrink to one page without anybody feeling short-changed, because the evidence moved somewhere more honest.
The bad month is the real problem
An honest report has to be able to say the month was bad. Almost none of them can, and that isn’t a writing failure.
Think about what the monthly cycle actually is from the other side. Twelve occasions a year on which somebody has to justify their fee, with no context in which being wrong is normal and no unit of time long enough for a loss to be part of a plan. I argued separately that a retainer is re-sold silently every month in the client’s head, and the monthly report is that re-selling turned into a document. Under those conditions, writing “we tried something and it didn’t work” is a professional risk, and people respond to incentives.
So if you want honest reporting, the thing to change is not the template. It is to say out loud, early, that you expect some months to be bad and you would rather hear about one than be managed through it. That sentence costs you nothing and it is the only thing that makes the rest possible.
Then notice whether anything changes. An agency that still can’t tell you a month went badly after you’ve explicitly invited it is telling you something more useful than any chart.
What I’d ask for next month
Ask for one page: what we said, what we did, what it cost, and what we’re changing.
Ask for Admin on your own accounts, this week, while everyone is friendly.
And tell whoever writes your report that a bad month, reported early with a changed plan attached, is worth more to you than a good-looking one. Then see if the next document gets shorter, which is the tell either way.
If you want a second opinion on whether your reporting is telling you anything or just reassuring you, drop me a line on email, WhatsApp, phone or LinkedIn and we can have a quick chat, and my notes on how I work cover the rest. I’m taking on a small number of contracts at the moment, so I’ll tell you straight whether it’s something I could help with.