Setting a Marketing Budget That Scales With Revenue

Most companies land somewhere between 6 and 13 percent of revenue on marketing, and the share tends to fall as the company gets bigger. The 2026 CMO Survey puts the average at 8.96 percent across 154 US firms. It also reports a standard deviation of 11.45 points, which tells you the average alone is nearly useless as a planning input. A budget that genuinely scales with revenue is not one percentage. It is a floor you fund whatever happens, a variable layer priced from your own acquisition math, and a short list of written triggers that move the number when revenue moves. This guide shows how to build all three.

What the Benchmarks Actually Say

Start with the published numbers, because they are the first thing a board or a CFO will reach for.

The CMO Survey, run by Duke University's Fuqua School of Business with Deloitte and the American Marketing Association, surveyed 308 US marketing leaders in January 2026. Marketing came in at 9.0 percent of company revenue and 9.6 percent of total company budgets, both multi-year lows. Overall marketing spending rose 1.7 percent over the prior twelve months, the smallest increase since 2021.

Cut that by company size and the pattern gets more interesting. The share does not climb as a company grows. It drops, and it drops hard somewhere past the 100 million mark.

Source: The CMO Survey, Firm and Industry Breakout Report 2026, p. 127.

Sector matters too. The same report has B2C product companies at 12.0 percent, B2B services at 10.1 percent, B2C services at 7.2 percent, and B2B product firms at 7.0 percent. Gartner's 2026 CMO Spend Survey, which polled 401 senior marketers at companies above a billion dollars in revenue, landed at 7.8 percent overall.

Two caveats before you anchor on any of these. The size bands rest on 15 to 34 responses each, so treat them as direction rather than truth. And in most bands the standard deviation runs larger than the mean, meaning the companies inside a single band disagree with each other more than the bands disagree with one another. Benchmarks tell you whether you are strange. They cannot tell you whether you are right.

Why a Flat Percentage Breaks

Pick a percentage and hold it, and three problems show up within a year.

The first is circularity. Revenue pays for marketing, and marketing is supposed to produce revenue. Fixing spend as a share of last year's sales means this year's growth gets funded by last year's results, which is backwards in a good year and brutal in a bad one.

The second is that costs do not scale in a straight line. Some of your marketing spend is fixed: the platform licences, the site, the people, the reporting. The rest varies with volume: media, production, events, partner fees. A single percentage treats both as if they flex together. They do not. Double revenue and your media line can double, while your analytics stack barely moves.

The third is averages hiding everything. The CMO Survey's 163 respondents moved overall marketing spend by a mean of 1.7 percent, with a standard deviation of 15.4 points. Half the market cut while half added. Nobody actually spent the average.

The Three Layers

A budget that scales needs three pieces, and they behave differently.

• The floor is what you spend to stay in business if growth stops tomorrow. Brand site, core tooling, the people who hold institutional knowledge, the measurement you need to see anything at all. Cut into the floor and you stop being able to run marketing, not just stop growing.
• The variable layer is acquisition. Media, production, outbound tooling, anything whose output rises when you feed it more money. This layer is priced from unit economics, not from a percentage.
• The bet layer is the small slice you set aside for things with no proven return yet. A new channel, a category you want to own in AI search, a conference nobody in your market attends. Keep it modest and keep it named, because an unnamed bet layer gets raided the first time a quarter goes sideways.

The Five Steps

Step 1: Set the floor in dollars, not percent

Add up what you would still pay next year at zero growth. Software, salaries, retainers you cannot exit, hosting, the analytics contract. That number is a dollar figure and it should barely move with revenue. Write it down separately, because it is the part of the budget that has no business being expressed as a percentage.

Step 2: Price the variable layer from your own acquisition math

Take your three inputs: what a customer is worth over their life, what you currently pay to acquire one, and how long the payback takes. If a customer is worth 9,000 dollars in gross profit and costs 3,000 to acquire, each acquisition dollar buys three dollars of margin, and your constraint is cash, not appetite. Decide how many new customers the revenue plan needs, multiply by your acquisition cost, add the production needed to support it, and you have a variable layer that falls out of the plan rather than out of a benchmark. If you cannot state your acquisition cost with a straight face, that is the first problem to fix, and our guide to the KPIs that actually predict revenue is the place to start.

Step 3: Write the triggers before you need them

This is the step that makes a budget scale, and almost nobody does it. Agree in writing, now, what changes the number later. A trigger has three parts: the condition, the action, and the owner.

• If trailing three-month revenue beats plan by 10 percent and acquisition cost holds flat, the variable layer rises by an agreed amount next month.
• If acquisition cost rises more than 25 percent over a quarter, new channel spend pauses until the cause is identified.
• If revenue misses plan by 10 percent for two consecutive months, the bet layer goes first, then the lowest-performing paid channel, in that order.

Written triggers convert budget arguments into arithmetic. They also stop the worst pattern in marketing finance, which is cutting whatever is easiest to cancel in the week the news arrives. The ownership half of that sits inside the marketing accountability framework.

Step 4: Decide which revenue number you are scaling against

Trailing revenue is safe and slow. It funds growth you have already banked and leaves you underfunded in a fast quarter. Forward revenue, meaning the plan, funds the growth you want and overcommits you if the plan misses. Most companies between 2 and 50 million dollars do best with a blend: fund the floor and roughly 80 percent of the variable layer off trailing revenue, and release the rest against the forward plan as the quarter proves itself. Say which one you chose out loud. Half the budget fights inside companies are two people using different denominators.

Step 5: Re-forecast quarterly

An annual budget set in November is a guess about a year you have not seen. Keep the annual envelope, then re-forecast the variable layer every quarter against the triggers, with the same four or five numbers each time. Our annual planning guide covers the yearly build in detail; this is the lighter cadence that sits on top of it.

Scaling Down Is Part of the Design

A budget that scales with revenue has to scale in both directions, and most companies only design the upside.

The CMO Survey found that when year-to-date profits come in below expectations, executives focus on cutting expenses 53.1 percent of the time, up from 46 percent the year before. Marketing gets cut in 45.4 percent of those cases. Those are decent odds that someone will come for your budget this year, so decide the order now, while nothing is on fire.

A defensible cut order usually runs: the bet layer, then the weakest-performing variable channel, then production volume, then the floor, which you touch last and reluctantly. Write that sequence into the budget document. When the call comes, you are presenting a plan rather than negotiating under pressure, and the difference shows. Deciding which channel is weakest is its own exercise, and we walk through it in How to Tell If Your Marketing Spend Is Working.

The Arithmetic, End to End

Here is the shape of it, with illustrative numbers rather than a real client.

A B2B services business at 8 million dollars in revenue plans 11 million for next year. Its floor comes to 240,000 dollars: two salaries, the stack, the site, a PR retainer. Acquisition cost sits at 4,500 dollars per new client, and the plan needs 60 net new clients, so the variable layer prices out at 270,000 plus 60,000 of production support. The bet layer gets 40,000, earmarked for AI search visibility. Total: 610,000 dollars, which is 7.6 percent of current revenue and 5.5 percent of the target.

Against the CMO Survey's 10.1 percent for B2B services, that budget sits low, which is a conversation worth having rather than an error. Maybe the plan is underfunded. Maybe the business converts better than its peers. The point is that the benchmark is now the last check, not the starting assumption, and the budget can be defended line by line.

Where This Goes Wrong

Three failure patterns show up again and again.

Agency fees get counted as working spend. If 40 percent of the media line is management fees, your real acquisition budget is smaller than the spreadsheet claims. Separate the two on the page.

The percentage becomes the goal. Once a target share of revenue becomes the metric, teams start spending to hit it rather than to produce customers. The percentage is an output of the plan, never its objective.

Nobody owns the number. A budget with five stakeholders and no owner drifts, and the triggers never fire because firing them is somebody else's job. One name, one number. If you are not sure who owns yours, a marketing audit will surface that faster than another planning meeting.

The Bottom Line

The honest answer to what percentage of revenue marketing should take is that the percentage is a check, not a method. Build the floor in dollars. Price the variable layer off acquisition cost and customer value. Keep a named bet layer. Write the triggers that move the number in both directions, pick your denominator, and re-forecast quarterly. Then hold the result up against the published benchmarks and ask whether you look unusual, and whether you have a reason. That is a budget that scales with revenue. A fixed percentage only looks like one.