The most common mistake on this topic is treating the answer as a single number. An owner finds a rule of thumb, applies it to last year's revenue, and calls the budget done. That figure was averaged across businesses with different margins, different sales cycles and different growth goals, so it fits almost nobody. It also assumes your marketing should look like everyone else's.
Turn the question around. Instead of asking what share of revenue is normal, ask what you need marketing to accomplish, what each customer is worth to you, and how much you can spend before it stops paying back. The share of revenue comes out of that work; it is not where you start.
Why the benchmark misleads
A benchmark blends together companies that are launching, holding steady and fighting for market share. A new shop with no regulars needs to spend heavily just to be found. A business with a long waiting list needs far less. With the blended number, the first business underspends and the second wastes money, and neither can tell which has happened.
Revenue is also a poor base, because it ignores how much of each sale is left after costs. Two businesses with the same sales can have very different room to spend if one keeps a thin margin and the other keeps a fat one.
Start from margin and customer value
Work out what an average new customer is worth to you over the time they stay, then subtract the direct cost of serving them. What remains is the most you could ever pay to win them. Your real ceiling is lower, because you need profit left over, but this number is the hard boundary. Any budget you set should sit well inside it.
What should push your budget up or down
| Your situation | Direction for spend | Reason |
|---|---|---|
| Brand new, few customers know you | Higher | You need to be discovered before anything else works |
| Steady repeat business and referrals | Lower | Existing customers already do much of the selling |
| Wide margins on each sale | Higher | You can afford to pay more to win a customer |
| Thin margins | Lower, and tightly tracked | Little room for wasted spend |
| Want fast growth | Higher | Growth costs more than holding steady |
| Slow, seasonal demand | Time it to the season | Spending in the off-months rarely pays back |
Interest in this question is small and uneven
Searches for this question are modest: the whole cluster of 17 related queries adds up to about 40 a month. For the exact phrase, interest swung from a low of 10 a month in 2025-10 to a high of 140 a month in 2026-07. That fits what owners tell us: the question tends to surface at budget time, or when sales dip and someone wonders whether they are underspending. Don't wait for that moment. Set the budget when you are calm, not when you are worried.
Build the budget in three steps
Decide the goal in plain terms, such as more new customers, more repeat orders or a fuller calendar in a slow season. Next, estimate how many customers you can realistically serve, so you do not buy demand you cannot fill. Then fund a test rather than a whole year: spend a modest amount on the channel most likely to reach your buyers, and watch what comes back before committing more.
Keep free and cheap work in the plan too: your business listing, asking happy customers for reviews, and emailing people who already bought. These often return more per dollar than paid channels. For more on stretching a small budget, see our guide on setting a good advertising budget for a small business.
Measure before you raise the number
A budget is only as useful as your ability to see what it produced. Make sure calls, form fills and purchases are recorded and tied to where they came from. Then compare what you spent with what those customers brought in. Our guide to calculating ROAS shows how to do that without flattering yourself. If the return holds up, raise spend gradually. If it does not, fix the offer or the tracking before adding money.
Where outside help fits
You can do all of this yourself, and many owners do. Help is worth considering when the budget is large enough that mistakes are expensive, when your tracking is unreliable and you cannot trust the numbers, or when you simply lack the hours. A good outside partner should start by checking how results are measured, give you a clear picture of what is working, and let you see your own account. If someone offers a percentage without asking about your margins, treat that as a warning sign.
Whatever you decide, the next step is the same: write down your customer value, choose one goal, and set a test budget you could lose without losing sleep.