Somewhere between 4% and 20% of revenue, depending on your age, margins and who you sell to. That is the truthful range hiding behind every guide about how much a small business should spend on marketing: an established company defending a steady position can run on 4–7% of annual revenue, a business that wants real growth should plan for 7–12%, and a young company still proving its offer usually needs 10–20% of projected revenue to get anywhere at all. B2B sits toward the bottom of each range, consumer brands toward the top.
Stop reading here and you'll be roughly right. The problem is that "percent of revenue" is only half an answer, and the missing half, what the percentage assumes and where it stops working, is exactly where small budgets go to die.
Where the "7–8% of revenue" rule comes from
Almost every guide repeats the same figure: spend 7–8% of gross revenue. It traces back to old US Small Business Administration guidance for companies under USD 5 million in revenue, and it carried an assumption nobody quotes: net margins of 10–12% after all expenses. Hand that same 8% to a consultancy running 60% gross margins and to a trading company squeezing out 12%, and you've given one of them sensible advice and the other a cash-flow problem.
The big surveys don't transfer cleanly either. Gartner's 2026 CMO Spend Survey puts marketing at 7.8% of company revenue, barely moved from 7.7% the year before. Useful context, except the 401 companies surveyed mostly report revenue above USD 1 billion. The CMO Survey, run by Duke's Christine Moorman with Deloitte, reported budgets at 9.4% of revenue in spring 2025, and its spread matters far more than its average: B2B product companies at 6.4%, B2B services at 9.0%, consumer product companies at 15.5%.
That spread is the one benchmark finding worth internalising. Consumer brands buy attention constantly; B2B firms spend less on media and more time in longer sales cycles. Copying an average across that divide tells you nothing.
Benchmarks describe other people's averages. Your budget has to describe a plan.
Why percentage rules break down for a small marketing budget
Percentage math quietly falls apart below a certain revenue. Take a Dubai business doing AED 1 million a year. Eight percent is AED 80,000, about AED 6,700 a month. As of mid-2026 that buys either a decent content retainer or a modest ad budget with someone competent managing it, not both. We priced this out in detail in our breakdown of what social media management costs in Dubai; the arithmetic is unforgiving.
Channels also have minimum workable spends. Conversion-based Google Ads campaigns need enough data to learn, and as we covered in our UAE cost-per-click breakdown, budgets under roughly AED 3,000 a month in Dubai's auctions may never exit the learning phase. Meta is more forgiving, but a few hundred dirhams sprinkled across boosted posts is a donation, not a campaign.
So here's an opinion the benchmark articles won't give you: below roughly AED 5,000 a month, don't diversify. Pick the single channel closest to your customer's actual buying moment and fund it properly, even if the resulting number looks "overweight" as a percentage of your revenue. A concentrated budget that clears a channel's minimum beats a balanced budget where every channel is underfed.
The reverse is also true, and just as ignored. If you can prove a channel brings in customers for less than they're worth in margin, capping it at a tidy percentage of last year's revenue is self-sabotage. Percentages are for planning under uncertainty. Once the maths works, the maths is the budget.
Setting the number by stage, not by survey
When we build budgets with clients, the useful question is never "what does the average company spend?" It's "what job is marketing doing for you this year?" That gives three honest situations.
Proving. New business, new market, or a new offer inside an old business. Base the budget on projected revenue, 10–20%, and accept that this stage consumes cash before it returns any; that is what the stage is for. If that number is genuinely unaffordable, the answer isn't to run ads at a tenth of the workable spend. It's founder-led content, referrals and fast WhatsApp follow-up until there's cash to do paid properly.
Growing. The offer converts, and the question is throughput. Plan 7–12% of trailing revenue, positioned within the range by gross margin: high-margin services can justify the top, thin-margin trading businesses belong near the bottom. This is also the stage where the split between fixed and performance spending starts to matter more than the total.
Defending. Established position, stable pipeline, growth is nice rather than urgent. 4–7% maintains visibility, keeps the website and CRM from decaying, and funds enough presence that a hungrier competitor can't quietly take your ground. One warning from experience: businesses at this stage often cut to near zero in slow months, and in the UAE that usually means summer. Auctions get cheaper precisely when everyone else pauses; the discipline to keep spending modestly through Ramadan CPM spikes and August emptiness is what makes the January numbers look effortless.
Inside whichever range applies, four things should pull you up or down: gross margin, what a customer is worth over their lifetime, how expensive your competitive auction is, and whether you sell to businesses or consumers.
Fees are not marketing. Media is not marketing. Both are.
Whatever the total, split it deliberately into foundations and fuel. Foundations: a website that converts, tracking that works, a CRM that answers leads in minutes, a base of content. Fuel: media spend and the hands managing it. Early on the split leans toward foundations; in growth mode it flips, because by then every foundation dirham is amplifying paid traffic you're already buying.
The most common budget failure we see in Dubai isn't the total being wrong. It's an AED 10,000 monthly budget where AED 10,000 goes to a retainer and nothing is left for media, so a capable team spends a year producing content nobody sees. As a rule of thumb, once you're past the proving stage, keep at least a third of the total for media. If a proposal consumes your whole budget in fees, ask what, exactly, will carry the work to strangers.
An aside worth making: you are reading this on an agency's website, and agencies benefit when the answer to "how much?" is "more." Keep that in mind, and hold us to the same standard we're recommending: every dirham of fee should have a visible line to enquiries, not just activity.
When to cut, and when to double down
Cut when you can't see. If you don't know which channel produced last month's enquiries, spending more is gambling; fixing measurement costs less than one month of most retainers and changes every decision after it.
Cut when a channel has had a fair test and failed. Fair means the channel's minimum spend, decent creative, and roughly 90 days. Then kill it completely rather than shaving 20% off everything, which is the most popular and least effective way to reduce a budget. Feed the strongest channel; starve the weakest to death, not to weakness.
Double down when the payback maths is proven, meaning your CRM ties customers back to campaigns and cost per customer sits comfortably inside gross margin. At that point stop asking what percentage other companies spend. The honest ceiling becomes operational: how many new customers your team can actually serve. That's the point at which performance marketing stops being a cost line and starts being a throttle you set on purpose.
If you do one thing this week, don't set a percentage. Work out what one new customer is worth to you in gross margin, then check whether your current spend could possibly acquire customers below that number once fees are subtracted. If you'd like a second pair of eyes on that arithmetic, talk to us and bring your real numbers; the conversation is more useful than any benchmark.