How much should a service business actually spend on marketing?

The percentage-of-revenue rule everyone quotes is close to useless for a service business. Here is the number that actually decides your budget, and how to find it.

August 20, 2026 · 5 min read · by Ido Katz

Ask this question anywhere and you get the same answer: five to ten percent of revenue. Some people say seven. Some say twelve if you are growing.

It sounds authoritative, and it is close to useless.

The percentage rule was built for businesses selling products at a known margin to a broadly similar customer. A service business does not work like that. Two roofers on the same street, doing the same revenue, can have completely different correct budgets — because one gets half his work from referrals and the other gets none, or because one sells $1,200 repairs and the other sells $18,000 replacements.

A percentage of revenue tells you what other people spend. It tells you nothing about what you should spend.

The number that actually decides it

There is only one figure that settles this, and it is not a percentage. It is:

What you can afford to pay to acquire one customer.

Everything else follows from that. Work it out and the budget question answers itself; skip it and you are guessing no matter how sophisticated the spreadsheet looks.

You need two things you already have.

What a customer is worth to you over twelve months. Not the first invoice — everything they spend in a year, including repeat work. This is the number owners get wrong most often, almost always downward, which makes their advertising look far more expensive than it is.

What share of that you are willing to give up to win them. For most healthy service businesses this lands somewhere between a fifth and a third of first-year value.

Multiply. That is your maximum cost per customer. There is fuller arithmetic here, including how close rate fits in.

Turning that into a monthly budget

Now the budget is arithmetic rather than opinion:

Customers you want per month × what you can pay to acquire one = your monthly budget.

Say a customer is worth $2,400 in year one and you are willing to spend a quarter of that — $600 — to win one. You want eight new customers a month.

8 × $600 = $4,800 a month

That is your number. It has nothing to do with your revenue and everything to do with what you want to happen next month.

Notice what this does: it turns marketing from a cost you tolerate into a quantity you buy. If eight customers is worth $4,800 to you, twelve is worth $7,200, and the only real question is whether you can serve twelve.

Why the percentage rule survives anyway

Because it is easy, and because it protects people from the two failure modes at either end.

A business spending nothing has no idea whether marketing works, because it has never run it long enough to find out. A business spending wildly, with no cost-per-customer figure, can burn a quarter's profit before anyone notices the jobs did not follow.

The percentage keeps you out of both ditches. It just cannot tell you the right answer, and it will be badly wrong in both directions for perfectly ordinary businesses.

The floor nobody mentions

There is a practical minimum, and it has nothing to do with your size.

Below roughly $2,000 a month on paid advertising, the platforms never gather enough data to get efficient. Meta and Google both need a certain volume of conversions before their optimisation does anything useful. Under that threshold you are paying for the expensive part — the learning — and leaving before the part you were paying for arrives.

This is the single most common way service businesses conclude that "ads do not work". They spent $600 a month for three months, got a handful of leads at a terrible cost, and stopped. The ads did not fail. They never started.

If $2,000 a month is genuinely not available, do not split it across channels. Put all of it into one, and give it ninety days.

What the money should actually buy

A budget is not only media. A realistic split for a service business starting out:

Most businesses put 100% into the first row and wonder why the leads do not convert. The offer and the follow-up are not overhead — they multiply everything the media buys. A better offer costs nothing per lead and lifts every lead you ever get afterwards.

When to spend more

Spend more when — and only when — three things are true at once:

1. Your cost per customer is comfortably below what you can afford.
2. You can actually serve the extra work without service getting worse.
3. The last increase produced roughly proportional results.

That third one matters. Demand is not infinite in a local market. At some point the next dollar reaches people who are less likely to buy, and your cost per customer climbs. Raise budgets in steps of twenty or thirty percent and watch what happens to cost per customer, rather than doubling and hoping.

If cost per customer stays flat as you scale, keep going. If it climbs sharply, you have found the ceiling of your market at that offer — and the fix is a better offer, not a bigger budget.

When to spend less, or nothing

If your close rate is under about ten percent, more traffic is the wrong purchase. You have a filtering or offer problem, and buying more of the same mix makes your week worse without making you richer. Fix who you are quoting first.

If you cannot serve what you already have, stop. Creating demand you cannot meet costs you money twice — the ad spend, and the reputation.

The four minutes

1. Work out what a customer is worth to you over twelve months.
2. Decide what share of that you will pay to win one.
3. Decide how many new customers a month you actually want and can serve.
4. Multiply. That is your budget.

If the answer is under $2,000 a month, the honest advice is to wait until it is not, or to spend the time on referrals and repeat work instead — both of which cost nothing per customer and are usually underworked.

If the answer is well above $2,000 and you are currently spending a fraction of it, you do not have a budget problem. You have an unbought opportunity, and someone else in your market is buying it.