Start with a question nobody asks during budget season.
How many new customers could you actually serve next month without the work getting worse?
Not want. Serve. With the people you have, at the quality that keeps them. For most small businesses that number is lower than the growth target, and it is the only figure in this whole exercise that isn’t aspirational.
Which is why a percentage of revenue is the wrong shape of answer. It derives spending from what you have already sold, when the question is what you can deliver and what it costs to find somebody who wants it.
I went through why the percentage benchmarks describe billion-dollar companies maintaining demand they already have when writing about annual planning, so I won’t re-run it. One thing to add, since you will see it quoted: the “SBA recommends 7 to 8%” line appears on dozens of agency pages and I could not find an SBA page saying it. Treat it as folklore with a government logo attached.
The replacement is arithmetic, and it runs in an order most people reverse.
Capacity first, because it is the only real number
Work out your ceiling before you work out your ambition.
If you install something, how many installations can your crew complete in a month at the standard you advertise? If you sell services, how many new clients can be onboarded before onboarding becomes the bottleneck and existing clients notice? If you ship product, where does fulfilment start failing?
Then subtract what you already get without trying.
The answer is your actual headroom, and it does two things. It caps the budget, because there is no point buying demand you will serve badly. And it occasionally ends the conversation entirely.
If your ceiling is eight new customers a month and you are already getting six, you do not have a marketing budget problem. You have a capacity problem or a margin problem, and spending money to attract two more customers you will handle poorly buys you a worse reputation at a higher price. That is the most common misdiagnosis I see, and the budget conversation hides it because budgets are easier to talk about than capacity.
Then what a customer actually costs you, which you can find out by asking
This is where most people stop, because they do not know their acquisition cost and assume finding out requires software.
It does not. It requires twenty phone calls.
Take your last twenty customers and ask each one how they first heard about you. Not a survey link, an actual question, asked by a person. Write the answers down. Then take last quarter’s marketing spend, all of it including the things you forgot were marketing, and divide it by the number of customers the answers plausibly credit to it.
That is bad data. It is also your data, and it beats any benchmark by an order of magnitude, because no published benchmark exists for your business. I looked for a current small-business acquisition cost figure worth quoting and there isn’t one, which is most of why the percentage rule survives: it is the only number available to somebody who hasn’t measured their own.
Two things to accept about the answer. It will be a range rather than a figure, and a range is fine, because the decision you are making is roughly how much rather than exactly how much. And the attribution will be wrong in places. A customer who says “I just found you on Google” may have heard your name from a neighbour six weeks earlier. Live with it. You are not auditing, you are sizing.
If the word “customer” is doing too much work here because your sale takes months, use cost per qualified lead instead and settle what qualified means before you count anything, which I argued at length in the four numbers worth putting in a report.
The sum, and the thing it will get wrong
Headroom times acquisition cost. That is your budget.
Eight customers of headroom, $600 to acquire one through the channel that demonstrably works, so $4,800 a month, and now you are arguing about a real number instead of a percentage.
Here is what that sum gets wrong, and it is worth building in rather than discovering. The acquisition cost you just measured is the cost at your current volume, and it rises as you scale. The cheapest customers come first: the people already searching for you, the referrals, the ones at the bottom of the funnel. Doubling spend on a channel does not double customers, because you start paying for people who are less ready.
So treat today’s number as a floor rather than an average. In most accounts I’ve scaled, the cost per customer drifted up somewhere between a fifth and a half as spend doubled, and the honest version of your plan has that in it rather than discovering it in month three and calling the channel broken.
Also, pick one channel for this sum, not an average across three. A blended acquisition cost is a number that describes no decision you can make.
When the number is bigger than you can fund
It usually is. And what happens next is the actual mistake.
Honest arithmetic says $14,000 a month. You have $4,000. So you take the plan and scale it down proportionally: a bit of search, a bit of social, a bit of content, a bit of email, all at a quarter of what each needs.
That is the single worst available option, and it is the one most businesses pick because it looks like prudence. You have now funded five channels, every one of them below the level at which it produces anything, and at the end of the quarter you will have no result and no information either. Nothing was tested, so nothing was learned. Each channel has a floor below which the spend is a donation, which I worked through for paid search specifically and which applies in some form to all of them.
The correct move is less comfortable. Fund one channel properly, the one with your lowest honest acquisition cost, at or above its floor. Accept the smaller number of customers that buys. Then write down, in a sentence, the growth you are choosing not to pursue this quarter and why.
That last sentence matters more than it looks. A budget that silently fails to deliver a growth target poisons the next three planning cycles. A budget that states “this funds four of the eight customers a month we have room for, because we can afford one channel properly rather than four badly” is a decision, and decisions can be reviewed.
Where a percentage is actually right
To be fair to the rule, it has a job.
A mature business with stable demand, budgeting to hold its position rather than grow it, is doing maintenance. There the acquisition cost arithmetic has already been run, repeatedly, and the percentage is a control that stops the budget drifting while nobody is looking. That is a legitimate use and it describes a lot of companies.
The useful distinction: a percentage is a reasonable constraint and a poor method. Run the bottom-up number, then check it against a percentage of revenue to see whether it is sane. If your arithmetic says you should spend 40% of revenue acquiring customers, something upstream is broken, most likely your margin or your price, and no budget fixes that.
Use the percentage as the sanity check at the end. Just don’t use it as the answer, because it cannot see your capacity, your acquisition cost, or the one channel that works for you, and those three things are the entire question.
If you’d like help putting an honest number on what a customer currently costs you, drop me a line on email, WhatsApp, phone or LinkedIn and we can have a quick chat, and my notes on performance marketing cover what to do with the budget once you have one. 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.