How much should a small business spend on marketing?

The short answer

Most small businesses land between 5% and 10% of revenue, and the range is that wide because the right number depends on your margin and how old the business is. A five-year-old clinic with steady referrals can hold at the bottom of it. A business opening its second location, or one nobody has heard of yet, spends at the top and sometimes past it. Below roughly $500 a month you are not running marketing, you are running a website — which is a legitimate choice, but it should be a deliberate one.

Why the number you keep reading doesn't apply to you

The most quoted benchmark in marketing is Gartner's annual CMO Spend Survey, which put marketing at 7.8% of company revenue in 2026, essentially flat since 2022. It is a good number. It is also drawn from 401 senior marketers in North America and Europe, and the vast majority of those companies report annual revenue above one billion dollars.

A billion-dollar company spending 7.8% has a marketing department, an agency roster and a brand people already recognize. Its percentage is buying maintenance. Yours is buying the first thousand people who find out you exist. Those are different jobs at the same percentage, which is why copying the percentage without copying the situation produces a budget that feels responsible and does nothing.

The useful version of the question is not what percentage other people spend. It is what one customer is worth to you, what one costs to reach, and how many you need.

Percentage of revenue is the wrong denominator

Revenue is the number every benchmark uses and the number that tells you the least. Two businesses on $600,000 a year, one at a 6% net margin and one at 22%, have almost nothing in common as budget problems. The first has $36,000 of annual profit, so an 8% marketing budget is $48,000 — more than everything the business earned. The second has $132,000 and the same budget is a third of it.

Do the calculation on margin instead. Take your gross margin, because that is the money a new customer actually contributes before overhead, and decide what share of it you are willing to hand back to acquire the next one. That share is a business decision. It is not a benchmark and nobody outside your business can tell you what it should be.

If you want to know how your margin compares to your industry rather than to your instinct, Innovation, Science and Economic Development Canada publishes Financial Performance Data covering more than a thousand industries, built from real filings, with revenue and expense breakdowns and profit margins by revenue band. Dentists, beauty salons and real estate agents each have their own report. It is free, and it is a better starting point than any marketing article's rule of thumb.

What a customer is worth over the whole relationship

It is easy to underprice your own customers, because the number in front of you is the first transaction. The person who came in for a $180 cleaning is not worth $180. They are worth that cleaning, twice a year, for however many years they stay, plus the work that gets found along the way, plus the two people they send you.

The arithmetic is deliberately crude and still more useful than anything else on this page. Average transaction, times visits per year, times the number of years a typical customer stays, times your gross margin. A salon client at $140 who comes six times a year for three years at a 55% margin contributes about $1,386 in gross profit. A dental patient at $400 a year for eight years at 60% contributes about $1,920. Use your own numbers, and use the median rather than the mean — one outlier client will flatter an average badly.

Two things fall out of this immediately. Retention is worth more than acquisition, which is why the recall email and the rebooking text usually beat the ad. And a business with a long relationship can afford to pay much more for a customer than its competitor down the street who is thinking in single transactions.

Allowable acquisition cost: the one number worth deciding

Once you know the gross profit a customer contributes, set the most you are willing to spend to win one. That figure is your allowable acquisition cost, and having it turns every marketing decision from an argument into arithmetic.

Pick the share consciously. A third of first-year gross profit is a conservative place to start if you need the cash back quickly. Half of lifetime gross profit is aggressive and appropriate if you are funded, patient, and confident in your retention number. What matters is that you choose, write it down, and then measure against it — because the same $200 spent to acquire a customer is either fine or reckless depending entirely on which of those two you decided.

Then set a payback period alongside it. If a customer takes fourteen months to repay what you spent acquiring them, that budget is a financing decision as much as a marketing one, and it needs to be sized against your cash position rather than your ambition.

What a customer costs to reach

Work backwards from the ad market, because it prices attention honestly. Below are the going Google Ads rates in the US, pulled from Google Ads in August 2026, for the moment a customer is actively looking.

The right-hand column is an assumption, not a measurement — one booking per ten clicks is optimistic for a weak landing page and pessimistic for a strong one. Use your own number once you have one. But it makes the arithmetic concrete: twenty new patients a month bought purely through ads is roughly $2,700 in media for a dental practice, before anyone is paid to make the ads or answer the phone.

Now put that beside the allowable acquisition cost you just set. If a new patient is worth $1,900 in gross profit over eight years and you decided you would pay a third of the first year to win one, $135 is not expensive. If you sell a single $60 service that nobody repeats, it is ruinous. Same number, opposite conclusion, and the only thing that changed is that you did the other calculation first.

Then compare it to the channels where the click is free. The map result sitting directly above the ad costs nothing per visit. A review that arrives this week keeps working next year. A page answering what an implant costs earns visits for as long as it stays accurate. That is why the sequence below matters more than the total.

The order to spend in

Whatever the number is, the sequence is close to universal for a local business.

First, the things that are free to run and cheap to fix. The Google Business Profile with every service listed. Reviews arriving on a schedule instead of in bursts. Your name, address and phone matching everywhere. This is a few hundred dollars of work, or a few hours of your own, and it changes what happens to every visitor the rest of the budget buys.

Second, the pages that answer what people search. One page per service, written in the words customers use, with real numbers where you are willing to publish them. These take about 90 days to move in search, which is exactly why they are the second thing and not the fifth.

Third, the steady drumbeat. Posts, reels, an email that goes out monthly. This is where most owners either overspend on production or quietly stop after six weeks.

Fourth, and only then, paid. Ads amplify whatever is already there. Pointed at a finished profile and a page that answers the question, they are the fastest channel you have. Pointed at a thin page and four reviews, they are a way of paying full price for a bounce.

Keep the fee and the media in separate columns

The most common budgeting mistake we see has nothing to do with the total. It is a single line in a proposal that reads "$2,000/month, ads included" and cannot be taken apart afterwards.

Two numbers belong on separate lines. The fee is what you pay a person or a team to do the work. The media is what the platform charges to show it, and it should sit in an account in your own name, on your own card, where you can see every dollar. When they are merged you cannot tell whether a bad month was a bad campaign or a quiet one, and you cannot leave without losing the campaign history that took months to build.

Two related traps. Paying a percentage of ad spend pays your agency more when you spend more, which is a strange thing to design into a relationship. And there is a floor below which paid advertising is not really running: under roughly $1,000 a month, a campaign gathers too few conversions to optimize toward anything, and you are paying for a learning phase that never ends. Better to run one channel properly than three at a level where none of them can learn.

A monthly scorecard that fits on one page

Review the same six things on the same day every month. The point is not the dashboard. It is that a number you look at monthly gets managed and a number you look at annually gets excused.

Ask where each new customer came from and write the answer down. Attribution software will disagree with your front desk, and for a local business the front desk is usually closer to the truth.

Signs your budget is wrong

Too low: your marketing consists of posting when you remember, your profile has fewer than twenty reviews, and the last thing you published was in the spring. That is not a budget problem you can fix with a bigger ad spend — it is a production problem.

Too high: you are running ads to a site you would not send a friend to, or paying a retainer where nobody can tell you how many hours a month go into your account. Both are common and both are fixable this month.

About right: something ships every week, you know what it cost, and you can name the number you are trying to move. That last part matters more than the percentage.

How we price it

Four lines, each sold on its own, Canadian dollars, and taking more than one earns no discount. Organic Growth is $1,000 a month and covers ten designed posts, five reels, five search-built articles and a newsletter, plus the website, reviews and local search. Paid Growth is $1,000 a month with the ad budget separate and in your own account, which realistically needs about $1,000 a month of media and covers spend up to $10,000. B2B Sales is $2,000 to set up and $1,000 a month. Automation starts at $1,500 to build and $500 a month.

The constraint that comes with them: search takes about 90 days to move. Paid is the only line built to shift a number inside a month; on the others the first month is work you can look at, not a result you can bank.

The reason those figures are on the page is that a budget conversation is impossible when one side will not name one. If a single line covers what you need, we will say so.