Marketing Budget Allocation Starts With the 24.5% Nobody Puts in the Deck

A marketing budget allocation framework decides how spend is divided across people, tools, and media before anyone argues about channels. Gartner's 2026 CMO Spend Survey puts the average budget at 7.8% of revenue, with labor alone taking 24.5% of it. The split that matters is fixed versus working dollars, and most teams set it last, which is why channel debates feel like fighting over leftovers.

I think most budget conversations start in the wrong place. Someone opens a spreadsheet, lists Google, Meta, LinkedIn, events, content, and starts assigning percentages. Three meetings later the plan is mostly last year's numbers with a 5% haircut, and nobody can quite say why paid search got 31% instead of 26%.

The teams that seem to handle this well (based on what the big benchmark surveys show, plus a lot of planning docs I've seen shared around) make three decisions in order: how much in total, how much is already spoken for, and only then how the working money gets split. That's the framework below. It's mostly just a sequence, honestly.

Pass one: the total is a stage decision, not a benchmark lookup

Everyone wants the magic percentage. The honest answer is that the benchmarks disagree, and they disagree for a reason.

Gartner's figure is 7.8% of company revenue for 2026, up from 7.7% the year before, according to Yahoo Finance's coverage of the survey. But that sample is 401 marketing leaders, and most of them work at companies with more than $1 billion in annual revenue. The CMO Survey from Duke, Deloitte, and the AMA lands higher, at about 9.0% of revenue, per Portada's write-up of the 2026 edition, with overall spending growth of just 1.7%. Different sample, different number. Both are averages of companies that probably don't look like yours.

The small-business rule of thumb is lower still. The often-repeated SBA guidance, as the Houston Chronicle summarizes it, is 7% to 8% of gross revenue, but only if you're under $5 million in sales and running a 10% to 12% net margin. That condition gets dropped every time someone quotes it, which is kind of the whole point of the condition.

So here's how I'd actually set the total:

• Maintaining share in a stable business: 5% to 8% of revenue. You're defending, not taking.
• Growing on purpose: 10% to 12%. Michael Tasner's 70/20/10 piece in Entrepreneur makes a good case for 10% of projected sales, not last year's revenue, because "you can't market into the past."
• Early-stage or entering a new category: 15%+ and budgeted as an investment, with the board knowing that payback is a 12 to 24 month question.

Benchmark for this week: take your current budget, divide by trailing 12-month revenue, and write the number down. If it's below 7% and leadership is asking for growth targets above 20%, that's the conversation to have before any channel split. The plan is underfunded and no allocation trick fixes that.

Pass two: subtract the money that's already gone

This is the step almost every allocation template skips, and it's where a lot of "we have no budget" frustration really comes from.

A marketing budget is really two pools: fixed costs (people, tools, agency retainers) and working media. Gartner's numbers show the fixed side growing: labor rose from 21.9% to 24.5% of the marketing budget between 2025 and 2026, according to the ANA's summary of the survey. Martech sat at 19.4% in the same survey, per MarketScale's breakdown, down from 26.6% in 2021.

Add those together and roughly 44 cents of every marketing dollar at a big company is gone before a single ad runs. Then agency fees come off the top too.

Think of it like a household budget where rent got quietly more expensive. You can argue about restaurants versus groceries all you want, but if rent went from 22% to 25% of take-home pay, the restaurant budget shrank whether you voted on it or not.

The practical move: build the budget as two sheets. Sheet one is committed spend (salaries, contracts, retainers, tool renewals with dates). Sheet two is working spend, which is whatever's left. Only sheet two gets split by channel. If you haven't looked at your tool contracts in a while, the renewal-date martech audit is usually the fastest way to move money from sheet one back to sheet two.

A rough ratio worth checking: if working media is under 50% of your total marketing budget and you're a performance-driven business, the fixed side has probably crept. Not always wrong (a content-heavy team runs on people, not media), but worth a deliberate decision instead of drift.

Pass three: split the working money by certainty, then by time horizon

Now the channel conversation. Two lenses, used together.

Certainty: the 70/20/10 tiers

The 70/20/10 model sorts channels by how sure you are they work. Seventy percent goes to proven channels with a clear line from spend to customers. Twenty percent goes to promising bets you're trying to promote into the proven tier. Ten percent goes to real experiments. Tasner's version adds the part that actually makes it work: rebalance every quarter, promoting what's working and demoting what isn't.

The weakness is that "proven" usually means "proven by platform-reported ROAS," and those numbers flatter everyone. Before a channel earns a spot in the 70%, I'd want at least one outside check on it. The guide on why every platform says your ROAS is great covers how to get a number you can trust, and a simple geo holdout costs close to nothing to run.

Time horizon: brand versus activation

The second lens is how long you're willing to wait for results. Les Binet and Peter Field's work for the IPA famously landed on roughly 60% brand building and 40% short-term activation for consumer brands. For B2B, their analysis for LinkedIn's B2B Institute suggested about 46% brand and 54% activation, as this summary of The 5 Principles of Growth in B2B Marketing lays out.

Worth knowing the caveats, though. That B2B figure came from fewer than 50 cases, mostly large-budget UK campaigns, and the authors themselves called the conclusions "tentative." LinkedIn funded it, and LinkedIn sells brand advertising. None of that makes it wrong. It just means I'd treat 46/54 as a direction, not a target to hit to the decimal.

Personally, I'd adjust by stage. A company with almost no awareness and a short runway can't afford a 46% brand line. Something like 20/80 is more honest for a seed-stage team, moving toward 40/60 as you find product-market fit and the cost of each new customer starts climbing. That climb in acquisition cost is usually the signal that brand spend has started to pay for itself.

The retention line most budgets are quietly starving

This part surprised me when I lined the two big surveys up side by side.

Gartner found awareness and conversion now take 62.6% of total media spend, up more than 10% since 2024, while loyalty and retention spending fell 29% over the same stretch to under 15% of media spend (again via the ANA). Meanwhile, The CMO Survey found retention is now the strongest driver of marketing performance, ahead of both acquisition and brand. Christine Moorman, who runs the survey, called it "a notable reversal of the historical pattern in which acquisition and brand investment tended to lead."

So the thing that works best is getting the shrinking share. That's... not ideal.

Most budgets keep feeding the easiest part of the funnel to measure while the part earning the most gets a smaller slice every year.

My guess is this happens because acquisition channels are easy to automate and easy to report on, and retention work (lifecycle email, onboarding, customer marketing) mostly shows up as salary rather than media. It falls through the gap between sheet one and sheet two.

The fix I'd suggest is a retention floor. Set a minimum, maybe 15% to 20% of working spend for subscription or repeat-purchase businesses, that can't be raided to hit a quarterly acquisition target. If your repeat customers drive more than half of revenue and retention gets under 10% of working budget, that mismatch is probably the single biggest allocation fix available to you.

Putting the three passes on one page

Here's a worked example for a $20 million revenue B2B SaaS company that wants 25% growth:

• Total: 11% of revenue = $2.2M
• Committed: 4 marketers plus tools plus a content agency, call it $1.1M (50%)
• Working: $1.1M
• Retention floor: 15% of working = $165K for lifecycle, customer marketing, and expansion campaigns
• Remaining $935K split by horizon: roughly 35% brand ($327K) and 65% activation ($608K) at this stage
• Within activation, by certainty: 70% proven search and paid social, 20% a promising channel like Reddit or a partner program, 10% genuine experiments

Anyway, the specific numbers matter less than the order. If you change the total, everything downstream moves. If you skip pass two, you'll be splitting money you don't have.

One messy thought that doesn't fit neatly: none of this survives contact with a mid-year budget cut unless you decide in advance what gets cut first. Gartner reported that 56% of CMOs doubt they have the budget to execute their 2026 strategy. I'd write the cut order into the plan now (experiments first, then the 20% tier, retention floor last) so the decision isn't made in a panic in August.

Marketing budget allocation FAQ

What percentage of revenue should go to marketing?

Large companies average 7.8% (Gartner, 2026) to 9.0% (The CMO Survey, 2026). Profitable small businesses under $5M often use 7% to 8%. Companies actively chasing growth typically budget 10% to 12% of projected revenue, and early-stage companies often go higher.

What is the 70/20/10 rule for marketing budgets?

Seventy percent of working spend goes to proven channels, 20% to promising channels being validated, and 10% to experiments. Rebalance quarterly, promoting channels that prove out and cutting ones that decline.

How should B2B companies split brand and performance spend?

The B2B Institute's analysis suggests roughly 46% brand and 54% activation, based on a small sample of large-budget campaigns. Earlier-stage companies usually run closer to 20/80 and shift toward brand as acquisition costs rise.

Prediction, with numbers attached: I'd bet that by the 2027 survey cycle, at least one of the two major CMO surveys shows retention's share of media spend recovering above 15%, because the gap between where the money goes and where the returns are is just too visible now. Your own customer data will probably tell you the same thing a year earlier, if you put the retention line on the same page as everything else and actually look at it.