The marketing agency we paid for six months before anyone asked what it was doing

The pitch was good. It was good in the way that pitches from people who pitch for a living are good: confident, well-designed, full of examples from businesses that looked a bit like ours. We were busy, we knew our marketing was neglected, and here was a team who clearly knew more about it than we did. We signed at £1,800 a month on a twelve-month term and felt, for about a fortnight, like a company that had sorted something out.

Six months later I sat down to work out what we had got for £10,800 and could not answer the question. Not 'the answer was disappointing' — I genuinely could not tell. That is the part I have thought about most since, because it was not the agency's failure. It was ours.

Nobody wrote down the number

We never established a baseline. When they started, nobody in the business could have told you how many enquiries we received in an average month, where they came from, or what proportion turned into paid work. So when the reports arrived — and they arrived punctually, every month, beautifully laid out — there was nothing to compare them to.

The reports showed impressions, reach, engagement, follower growth and something called sentiment. All of it went up. None of it was the thing we were buying. We were buying more enquiries from the kind of customer we wanted, and the word 'enquiries' did not appear in six months of reporting.

When I finally counted properly, going back through the inbox and the call log, we had been getting somewhere between twenty and twenty-five enquiries a month before the agency started, and between twenty-two and twenty-eight after. On the most generous reading of that — three extra enquiries a month, converting at one in three, average job value £900 — the work was generating about £900 a month of extra revenue against £1,800 a month of cost. We were paying two pounds to make one.

If the monthly report does not contain the number of enquiries, it is not a report on your marketing. It is a report on the agency's activity.

Agencies report what they control

This is the honest structural problem, and it is worth understanding rather than resenting. An agency controls posts, ads, impressions and clicks. It does not control whether your phone gets answered, how fast you quote, whether your prices are competitive, or whether the person who called was ever going to buy. So it reports the things it controls, because those are the things it can be held to fairly.

The trouble is that a report full of things the supplier controls is a report about the supplier. It is not a management report. The one line that turns it into a management report is enquiries this month against the average of the previous three, split by source, with a cost per enquiry attached. That is one row of a spreadsheet, and any competent agency can produce it if you ask.

We did not ask, for a reason I suspect is common. We had spent years pushing back hard on quotes for materials, on van leases, on insurance renewals — and we never once pushed back on a marketing invoice, because we did not feel qualified to have the argument. You do not need to understand paid social to ask what happened to the enquiries. That is the whole lesson, really.

The meeting we should have had in week two

When we did eventually sit down for a proper review, the account manager was straightforwardly decent about it. She pulled the numbers we asked for within two days, and they were not flattering. And she said something that has stuck with me: nobody had ever asked her for that report before.

Our brief, when I went back to the email that started it all, had been 'we need more visibility'. We got visibility. It is very hard to be annoyed with a supplier who delivered precisely what was asked for by a client who asked for the wrong thing.

What we would insist on now

Agree the primary number before signing. One number, written into the first page of the proposal: qualified enquiries per month. A secondary number underneath it: cost per enquiry. Everything else is context.

Set the baseline yourself in the week before they start. Count last quarter's enquiries by source, however roughly, and email the figure to the agency so both sides are working from the same starting line. If you cannot count them, that is the first job — and it is yours, not theirs.

Buy a ninety-day test, not a twelve-month retainer. At £1,800 a month, ninety days is £5,400: a decision you can reverse, on a budget you can afford to be wrong about. Agencies push for twelve months because the first two are genuinely set-up and they do not want to be judged on them, which is fair — so concede a three-month minimum term with a break clause after it, rather than conceding a year.

Own the assets. The ad account, the analytics property, the domain, the Google Business Profile and any tracking should all sit in accounts registered to you, with the agency granted access as a user. This is the clause that costs people real money on exit, when the ad history and the pixel data that took a year to build turn out to belong to someone else. Your Google Business Profile: the free hour that brings in local work covers why that listing in particular should never sit in an agency's account.

Split the fee and the media spend on the invoice. A single figure hides which is which, and you cannot judge either without seeing both. And put the review meetings in the diary at day forty-five and day ninety when you sign, not when you start feeling uneasy — a review you booked in advance is a routine conversation, while one you call at month six is an ambush and everyone behaves accordingly.

What we did next

We ended it at month seven and kept one thing they had built, which was the tracking. Then we spent about a third of the old budget on the unglamorous local basics — the profile, the review habit, faster responses to enquiries that were already coming in — and measured what happened, using an actual baseline this time.

The result was better, though I want to be careful about the moral. It is not that agencies are a waste of money; plenty of businesses get excellent value from them. It is that a supplier can only be as good as the brief and the questions they get asked, and marketing is the one line of the P&L most owners feel unqualified to interrogate. Put the enquiry number next to the other figures you already look at every month — five numbers every UK business owner should know cold is a decent place to start — and the conversation stops being about marketing at all. It becomes a conversation about a supplier and a result, which is a conversation every owner already knows how to have. If you are setting a budget from scratch, how to set a marketing budget with no baseline works through the arithmetic.

Common questions

How much should a small business spend on marketing?

There is no percentage of turnover that answers this honestly, because the right number depends on your margin and how much a customer is worth over their lifetime rather than on your size. Work backwards instead. Decide what a new customer is worth in gross profit, decide how many enquiries it takes to win one, and you have a maximum sensible cost per enquiry. Then set a test budget you could afford to lose entirely — ninety days is usually enough to see a trend — and only scale spending once the cost per enquiry comes in under that ceiling with real numbers behind it.

Is a twelve-month agency contract normal?

It is common and it is negotiable. Agencies ask for twelve months because the first two months are largely set-up, research and learning, and they do not want to be judged on results that have not had time to appear. That argument is reasonable as far as it goes. The compromise most will accept is a three-month minimum term followed by a rolling arrangement with thirty days' notice, plus a written review at day forty-five and day ninety. If an agency will not discuss any break clause at all, treat that as information about how confident they are.

What should be in a monthly marketing report?

Four things, and they fit on one page. Enquiries this month, split by source, against the average of the previous three months. Cost per enquiry, calculated using fee plus media spend combined. What was actually done during the month, in plain language. And what will be done next month, with the reason. Impressions, reach and engagement can appear underneath as supporting detail, but they are not the report. If a supplier cannot produce the enquiry figure, the tracking is not set up properly, and that is the first thing to fix before spending another month.

Who should own the ad account and analytics property?

You should, without exception. Create the accounts in your own name or the company's, then invite the agency in as a user with admin rights. Ad accounts accumulate history and audience data that materially affect performance and cost, and analytics holds the record you need to judge anyone who comes after them. If an agency set things up in their own account, ask for ownership to be transferred rather than for a copy of the data — a data export is not the same asset. Do the same with the domain, the website hosting and the Google Business Profile.