How much should a startup spend on marketing in 2026? An honest answer.

The percentage-of-revenue model breaks for early-stage businesses

Ten per cent of £50,000 revenue is £5,000 a year. That's less than the cost of one month of meaningful paid advertising, let alone a brand, a website and a growth programme. The percentage model is a useful heuristic when your baseline is large enough to fund what's actually needed. Below that threshold, it doesn't give you a number that's actionable. It gives you a number that's too small to do anything meaningful with.

Before the percentage model becomes relevant, the right frame is different: what do I need to achieve in the next 90 days, what's the minimum viable investment to achieve it, and what's the realistic return if I get it right? That question produces a budget figure anchored to outcomes rather than a formula anchored to historical revenue, which is almost always the wrong anchor for a business that's growing.

This doesn't mean spend without discipline. It means the discipline has to come from clarity about what you're trying to achieve, not from a percentage applied to a number that may be entirely unrelated to the scale of marketing activity required.

The right sequence before the right budget

The biggest marketing budget mistake early-stage founders make isn't overspending. It's spending in the wrong order. Running Google Ads to a website built in 2019 on Wix. Running Facebook ads to a landing page that doesn't convert. Building brand awareness for a brand that isn't defined well enough to be recognisable in the first place.

The correct sequence is: brand foundation first, then website, then paid traffic, then content. In that order. Each layer multiplies the effectiveness of what comes after it. Out of sequence, each layer underperforms because the foundation beneath it isn't solid enough to support it. You can buy traffic to a weak site. But you'll pay twice: once for the traffic, once for the conversion rate that never materialises.

A well-defined brand makes a website easier to build and more coherent once it's live. A website that converts makes paid campaigns profitable at lower cost-per-click thresholds. Paid campaigns that are profitable provide the signal and the revenue to invest in longer-term content. Skip the sequence, and the budget required at each stage multiplies because you're compensating for what the previous layer failed to do.

What each stage of investment looks like

Concrete ranges are more useful than principles, so here are the numbers as they actually look in practice, broken down by stage, as management fees. Ad spend is separate, controlled by you, and scales with revenue.

Pre-launch and very early stage (sub-£50k revenue): the priority is foundation, not distribution. Invest in brand and website first, typically £5,000–£10,000 as a one-off, covering brand identity, positioning and a conversion-ready website. Add paid advertising only after those are in place. Running ads before the site is ready is the most common and most expensive sequencing mistake at this stage.

Early growth stage (£50k–£250k revenue): brand and website are done. The focus shifts to acquisition. A managed growth programme combining paid search and content typically costs £1,500–£3,000 per month in management fees. At this stage you're building the data and the pipeline that informs what you scale into next.

Scaling (£250k+ revenue): all channels running simultaneously, with budget allocated based on performance data rather than intuition. Management fees at this stage typically run £4,000–£8,000 per month, with ad spend layered on top and scaled according to what the campaigns are returning.

The jump between stages is intentional. You don't move to the next level of spend until the current level is producing a return worth building on.

The false economy of underspending

The most common startup marketing mistake isn't overspending on the wrong things. It's underspending to the point where nothing works well enough to produce a conclusion.

A Google Ads campaign with a £5 daily budget in a competitive London market will exhaust its spend by 9am, deliver no meaningful impression share, generate no optimisation data and produce no pipeline. A content programme with one article per quarter will rank for nothing, build no topical authority and drive no organic traffic. Marketing investment below the threshold required for effectiveness produces no return. It creates the false conclusion that marketing doesn't work. The real problem is that the investment was never large enough to find out.

Every marketing channel has a minimum effective dose. Below that threshold, you get noise. Above it, you get data. The data is what you make decisions from. Underspending means you never reach the threshold where the channel tells you whether it works. You just stop before you find out, conclude it didn't work, and move on to the next underinvested experiment.

If a budget genuinely can't reach the minimum effective threshold for a given channel, that channel shouldn't be running. Better to do one thing properly than three things at sub-threshold spend.

The question to ask yourself

Stop asking "how much should I spend on marketing?" It's the wrong question. It produces answers that are either too abstract (7–12% of revenue) or too arbitrary (what can I afford this month?).

Ask instead: what do I need to achieve in the next 12 months? What is the cost, in lost revenue, lost market position, lost momentum, of not achieving it? And how much would I need to invest in marketing to give myself a real chance of getting there?

Frame the marketing budget as an investment with an expected return and a required timeframe, not as a monthly operating cost to be minimised. The businesses that grow quickly treat marketing spend the same way they treat hiring: as a commitment to a specific outcome, with a minimum level of investment required to achieve it, and a clear view of what a return looks like.

That reframe, from cost to investment, is what separates founders who build scalable growth from founders who spend two years experimenting at sub-threshold levels and conclude that marketing is expensive and doesn't work.