What a marketing budget should actually look like

A percentage of revenue is a benchmark rather than a decision. It tells you what businesses of roughly your size are doing on average, across sectors with completely different customer values, margins and sales cycles, and it produces the same answer whether your customers are worth two hundred pounds once or twenty thousand pounds over five years.

It also fails in both directions. For a business with high customer value it produces a number far below what the economics could support, which caps growth for no reason. For a business with thin margins on one-off purchases it produces a number that cannot possibly be recovered, and the shortfall is discovered several months in. Neither failure is visible from the percentage itself, which is why the rule survives.

Four numbers produce the answer, and most businesses have three of them already. The worked figures below are illustrative arithmetic rather than any kind of benchmark — substitute your own, because the method is the point and the numbers are not.

1. 1.Average customer value. What one customer produces in revenue, over the whole relationship rather than the first transaction. A business whose customers stay three years has a customer worth three years, and budgeting against the first invoice is the most common self-imposed constraint there is.
2. 2.Gross margin on that. Revenue is not what you can spend from; the margin is. Apply your actual figure rather than an assumed one, and include the delivery cost of servicing that customer.
3. 3.What proportion of that margin you are willing to spend on acquisition. There is no correct answer — a growing business might accept most of the first year, a mature one considerably less. Deciding it deliberately is the point.
4. 4.Your close rate from enquiry to customer. If one enquiry in four becomes a customer, an enquiry is worth a quarter of what a customer is worth, and that is the number every channel gets measured against.

Suppose, purely as arithmetic, that a customer produces £6,000 over their lifetime at a 50% margin, that you will spend a third of that margin acquiring them, and that one enquiry in four converts. You can afford £1,000 to acquire a customer and therefore £250 for an enquiry. Every channel is now judged against £250, and the marketing budget is simply £1,000 multiplied by how many customers you want.

That final step is where most budgets are actually set, and it is a capacity question rather than a marketing one. Deciding you want thirty new customers when you can service twelve produces a budget that buys enquiries you will handle badly, which costs more than not having generated them.

Splitting by channel percentages is the same mistake at a smaller scale. The useful split is by what the money is doing, because those three jobs have different timescales and get judged differently.

The proportions depend entirely on where you are. A business with unserved existing demand should spend almost everything on capture until that runs out. One that has saturated search has no option but to create demand, which is slower and harder to attribute. And a business losing customers as fast as it acquires them should spend disproportionately on the third row, because acquisition onto a leaking base is buying the same customer twice.

Several things get charged to marketing that are not marketing, and they distort the figure enough to make every subsequent decision wrong.

• Agency fees counted separately from media spend. Your actual cost per customer includes both, and reporting media spend alone understates it, frequently by a third.
• The website rebuild. That is capital work with a multi-year life, and charging it to one year of marketing makes that year look catastrophic and the next look excellent.
• Sales commission. It is a cost of sale rather than a cost of acquisition, and mixing them makes both numbers unreadable.
• Anything that is really sales headcount. A business development hire is a sales cost, whatever the job title says.
• Platform fees for tools the whole business uses. A CRM is operations, not marketing, however much marketing depends on it.

Three situations where spending is the wrong move, and all three are common enough that any honest article on this subject has to name them.

The first is capacity. If you cannot service more customers, marketing spend produces enquiries you decline, which costs money and generates reviews from people you turned away. Raise prices instead — it is the only lever that improves margin without needing capacity you do not have.

The second is a leak downstream. If enquiries take three days to answer, or the close rate has quietly halved, more enquiries make the problem larger rather than smaller. Fixing response time typically costs nothing and improves the return on every pound spent afterwards.

The third is not knowing the numbers. A business that cannot state what a customer is worth cannot tell whether any channel is working, so the first spend is on measurement rather than on media. That is an unglamorous recommendation and it is the one that changes the most.