How much should a company spend on marketing? The big surveys put the average between about 8% and 9% of revenue: 7.8% in Gartner’s 2026 survey of chief marketing officers, about 8% for B2B firms in Forrester’s 2024 benchmark, and 9% in the 2026 CMO Survey. The CMO Survey’s median, though, was only 5%, so the typical company in it spends a good deal less than the average suggests.
I wouldn’t set your budget from any of those. They’re averages of what other companies spent, most of them much bigger than yours, and none of them knows your margins, your close rate or what next year’s goal asks of marketing. I’d build the number from your own economics, then test it two ways:
- against the goal, to see whether the budget can realistically deliver it
- against what the business can fund, to see whether the goal needs to move
Only then hold it up to the surveys to see whether something’s off.
What percentage of revenue do companies spend on marketing?
These three surveys are the ones worth knowing, along with who actually answered them.
| Source | Who answered | Figure |
|---|---|---|
| Gartner CMO Spend Survey, 2026 | 401 marketing leaders in North America, the UK and Europe, the vast majority at companies above $1 billion in revenue | 7.8% of revenue (7.7% in 2025) |
| The CMO Survey, 2026 | 308 U.S. marketing leaders, about a third at companies under $100 million; 154 answered this question | Average 9.0%, median 5%; B2B product 7.0%, B2B services 10.1% |
| Forrester, 2024 | Nearly 500 B2B organizations | About 8% for the average B2B firm |
They disagree for reasons that matter at your size. Gartner is mostly describing billion-dollar companies. The CMO Survey’s average is pulled up by a smaller group of heavy spenders, which is why its median sits four points lower. And the surveys don’t count the same things. Gartner’s budget covers marketing salaries, agencies, technology and media. The CMO Survey asks what share of revenue goes to marketing without defining every cost in the question, so companies that count salaries, trade shows or sales support differently can report different percentages for similar spending.
Company size pushes the other way. In the CMO Survey, companies between $10 million and $100 million reported spending a larger share of revenue than companies above $100 million. Those groups had only 15 to 18 respondents each, so I’d read it as a rough direction and wouldn’t plan on the figure. Part of the reason is probably that some costs, like a decent website and someone capable of running marketing, don’t shrink in proportion to revenue.
Does the SBA really recommend 7% to 8%?
You’ll often read that the U.S. Small Business Administration recommends spending 7% to 8% of revenue on marketing. The SBA’s own pages don’t appear to say that. The SBA page that ranks for the question is a 2019 guest post, and it says, “There’s no hard and fast answer to how much your marketing budget should be.” Its figure for marketing spend, 7.9% of revenue in 2018, came from a marketing agency’s summary of the CMO Survey. So the “SBA rule” looks like a survey average that picked up a government name along the way.
Build the number from what a customer is worth
The build has three parts: what you can spend to win each new customer, how many new customers the plan needs, and what marketing costs beyond winning new business.
- The most you can spend to win one customer. I’ve walked through how to work out what you can afford to spend to acquire a customer separately, with a calculator. It gives a ceiling for sales and marketing combined. Take out the cost of selling (salespeople’s time, commissions, travel, proposals) and what’s left is marketing’s share. In that post’s made-up example the combined ceiling is $15,000. Say $5,000 of it goes to selling; that leaves marketing $10,000. That’s what you can afford, not a forecast of what winning a customer will cost, so check it against what it has actually cost you.
- How many new customers the plan needs. Start from next year’s revenue goal, subtract what existing customers and dependable referrals will reasonably bring in (net of any customers you expect to lose), and divide the gap by what a new customer brings in during the planning year. Use that year’s revenue, not the annual or lifetime value: a customer won in September contributes only a few months. Say it comes to 16. Then check the pipeline can produce them: at your close rate, how many qualified opportunities 16 customers take, whether you’ve ever produced that many, and whether sales can handle them. That’s the subject of how many leads you need.
- What else marketing pays for. Keeping and growing the customers you already have (account marketing, client events, the newsletter): say $40,000. The base everything runs on (website upkeep, tools, the part of someone’s time that isn’t tied to new business): say $30,000.
At up to $10,000 for each of 16 customers, winning new business could cost as much as $160,000. Add the other $70,000 and this made-up company lands at about $230,000 for the year, with people, fees and media all inside that number.
That $160,000 is the most winning new business should cost, not a forecast. If winning a customer has actually cost you $12,000, those 16 customers take $192,000, the year comes to about $262,000, and each customer loses money, which is a problem to fix before it’s a budget to approve.
Hold it up against the surveys. On $10 million of revenue, $230,000 is 2.3%, well under every average in the table. That doesn’t automatically mean the plan is too thin. The surveys count things your budget may not, like whole marketing departments and brand campaigns, and a company that grows mostly through referrals and long relationships may genuinely need less. I once told an engineering services firm, in a strategy session, that its referrals were a steadier source of growth than the digital marketing my own agency sells. The first dollars there belonged to protecting and growing the referrals, with any digital spend judged against them.
One limit on all of this. The ceiling tests money meant to win customers inside the planning year. It’s the wrong test for long-horizon work like brand or organic content, where a year may be too soon to judge.
Goal first, or budget first?
There are two honest ways into this, and leadership teams can start from different ends. My own instinct is to start from the number: if you have $5,000 a month you’d do one channel as well as you can, and at $50,000 you can be effective in several, so the total tells you roughly what’s even worth attempting. Some leaders I work with ask the other way round: forget what we spend today, what would it take to hit the goal?
The plan needs both answered. Build the number from the goal, as above, then put it next to what the business can actually fund this year. If they meet, check that the leads, sales capacity and timing behind it are realistic. If they don’t meet, something has to give:
- a better close rate, a higher price or more value from each customer
- a cheaper way to reach buyers
- more budget, if the economics support it and someone is willing to defend it
- a smaller goal, or a longer timeline
If none of those will move far enough this year, it’s better to say so before the budget is set than to fund a plan the numbers never supported.
Check what the goal asks of the budget
When the budget is fixed first, run the test the other way. Treat the budget as set for a moment and ask how realistic the goal is. Take the revenue the goal needs from new customers and divide it by what’s left of the budget for winning them. That gives you the new-customer revenue every one of those dollars would have to bring in.
Then compare that with what your marketing has actually brought in before, over a comparable period and counted the same way. If the goal needs several times what you’ve ever seen, the plan is asking marketing for something it has no history of doing. Say so before the year starts, and decide what moves. It protects everyone: the marketing team isn’t set up to miss, and the sales team isn’t left blaming marketing for opportunities the budget could never have produced. It’s worth asking the downside question too: what happens if winning customers costs a quarter more than planned, or deals close a quarter later? The calculator below does the division for you.
Fund the base properly, and hold some back
A budget doesn’t have to be spent evenly across twelve months or spread across a quota of campaigns. I’d rather fund one channel properly as something that’s always on, then keep money set aside for the opportunities you can’t schedule: the event that has reliably produced pipeline before, the campaign that starts working and deserves more. Once something is producing proven results, it should be able to get more money, as long as the returns hold up as spend grows.
Keep a little contingency on top of that for the unexpected, like a new trade show or a video you suddenly need. Without a line for either, every good opportunity mid-year becomes an argument about where the money comes from. Once you have the total, here’s how to split it across channels.
People and fees vs media
Whatever the total, it falls into two broad categories. One is people and fees: your marketing team’s salaries, agency and freelancer fees, outside leadership, and the hours someone inside spends briefing vendors and reviewing their work, which cost money even though no invoice shows them. The other is media and tools: ad spend, sponsorships, events and software.
When a leadership team talks about “the marketing budget,” make sure everyone means the same one, and whether it includes sales costs like trade shows or sales support. In Gartner’s 2025 survey, paid media was 30.6% of the average marketing budget, and in its 2026 results, labor was 24.5%. Even at billion-dollar companies, most of the money isn’t advertising, so a budget judged only by its ad spend is being judged on less than a third of what those companies spend.
Ask how any outside help is paid while you’re at it. A fee that rises with your media budget gives whoever earns it a reason to recommend more media, so find out whether a fee tracks the work or the spend.
Plan the cash, and the review dates
If you sell something expensive or complex, marketing money goes out months before the revenue it wins comes in. The ceiling tells you what you can afford per customer; payback tells you how long you’ll wait for it. Check that the business can carry the spending through those months, and if the cash won’t stretch, phase the plan rather than cutting it off halfway through its first cycle. The same goes for new people: a new marketing hire or outside lead costs money before they start paying for themselves, so budget for that dip on purpose.
And don’t commit a full year to anything untested. For each channel or campaign, write down before the money goes out what you expect it to produce and at what cost, how long it runs before you look, and what you’d need to see on that date to keep going, change course or stop. Give it at least as long as your buyers typically take to reach a proposal, judge it on an early signal you agree in advance, and check that against closed sales once a full buying cycle has passed.
Your annual budget
Put in your own figures. The calculator builds the total, shows what it would be at your actual cost per customer, and, if a budget is already on the table, shows what the goal would need from it.
Build next year’s marketing budget
Start with the most marketing can spend to win one customer, then the number of new customers the plan needs. Add what it costs to keep the customers you have and the base everything runs on. If you know what winning a customer has actually cost you, add it too: the ceiling is what you can afford, not what it will cost. If a budget is already on the table, add it at the end to see what the goal would need from it. The examples in the boxes are made up.
For reference, the surveys in this post found a 5% median and 9% average (CMO Survey 2026), 7.8% (Gartner 2026) and about 8% for B2B firms (Forrester 2024), each counted differently. They’re a rough check, not a target. Revenue per dollar isn’t profit or return on investment. Nothing you type here leaves your browser.
Common questions about marketing budgets
How do I set a marketing budget if I don’t know my numbers?
Use your best estimate of what a customer is worth and what you’d spend to win one, label it as an estimate, and agree on how you’ll establish a baseline before committing to a bigger program. If you can’t tell where your good customers come from today, fixing that may be the first job.
Should a company spend more on marketing when sales are down?
Not automatically. First work out whether the problem is the number of qualified leads, how they’re followed up, or the offer itself. Spending more only helps if marketing is the bottleneck and the economics support it.
Do you need outside help with this?
Often not. If you have a marketing leader who already builds the budget from the economics and reviews it on agreed dates, you’re in good shape. If the plan is sound and nobody has time to carry it out, you need hands, and a good agency or a hire will do more for you than an advisor.
Outside help earns its fee when nobody owns marketing, or when you want the number tested by someone outside the building before you commit to it. Building that budget and the case for where it goes is a large part of my work as a marketing consultant. If you’d like to talk yours through, start on Work With Me.
