How much should a business spend on marketing, and how should it be split?

Ignore the percentages. The ones that get quoted come from surveys of billion-dollar companies, or from sources nobody can find. Work your budget out from your own numbers instead: how many customers you need, what each one is worth, and what you can afford to pay to win one. Then spend it on being found before being noticed, test small, and let the results set next quarter's split.

Why the usual percentages don't help

Two figures come up in almost every article on this question.

Gartner's 7.7% of revenue. Gartner's 2025 CMO Spend Survey found marketing budgets averaged 7.7% of company revenue, the same as the year before. It's a real figure from a real survey, but the survey was of 402 marketing leaders across North America, the UK and Europe, largely at companies with over $1 billion in annual revenue, as Marketing Brew and Chief Marketer reported. Half of those leaders had budgets of 6% of revenue or less. It tells you what large companies do, not what a firm of ten people should.

"The SBA recommends 7 to 8%." It's attributed to the US Small Business Administration all over the internet, but we couldn't find it in anything the SBA has published. Its own guide to marketing, before it was withdrawn, asked businesses to cost their plan line by line rather than giving them a percentage. Search results now point to a worksheet on another SBA site, which we haven't been able to check. Until somebody can show the figure at its source, treat it as unconfirmed.

The deeper problem with any percentage is that it starts from the wrong end. A budget should follow from what you're trying to achieve, not from what other companies happen to spend.

Work it out from your own numbers

You need four figures. You may have to estimate some of them the first time, and that's fine: an estimate you can improve beats a percentage borrowed from someone else.

1. How many new customers you need over the next year, and why that number.
2. What a customer is worth to you, in gross profit rather than revenue, over the time they typically stay with you.
3. How many enquiries it takes to win one customer, from your own sales records.
4. What you can afford to pay to win a customer and still make the margin you want.

After that, the arithmetic is simple. Here it is with round numbers:

• A business needs 40 new customers this year.
• Each customer is worth £2,000 in gross profit over the time they stay.
• It's prepared to spend up to a quarter of that to win one, which is £500.
• So the most it should spend on winning new customers is 40 × £500 = £20,000 over the year.
• If one enquiry in three becomes a customer, it needs 120 enquiries, so any channel costing more than about £167 an enquiry is working against it.

That last figure is the useful one. It turns "is Google Ads worth it?" or "is this agency any good?" into a question you can answer from the numbers each month.

If your figures produce a budget you can't afford, you've learned something before spending anything. Either the target is too high for now, the customer value is too low for paid channels, or the step from enquiry to customer needs fixing first. All three are cheaper to find out on paper.

What the budget has to cover

A marketing budget is more than advertising. Before you split it across channels, set aside the fixed costs that keep everything else working:

What's left after the fixed costs is what you have to spend on bringing customers in.

The order to spend it in

The order matters more than the percentages do, and it's the same order whoever your customers are.

1. Fix the foundations first

Before you spend on promotion, make sure the places people land are working. That means a website that explains clearly what you do, loads quickly on a phone and makes it obvious how to get in touch. The common problems are few and fixable. Sending paid traffic to a weak website pays for visitors who were never going to enquire.

It also means tracking that records enquiries properly. Without it, you can't tell which channel is working, and the rest of this plan is guesswork. You don't need much: five reports are enough to start.

2. Fund demand capture before demand creation

Demand capture means being there when people are already looking: search ads, SEO and your Google Business Profile. Demand creation means making people aware of a need they weren't thinking about: social ads, display, sponsorship, events. I think capture usually pays back faster, because you're reaching people who have already decided they want something, so fund it first. There's one exception you can check for yourself: if nobody searches for what you sell, there's nothing to capture, and creation is the only route open to you.

That doesn't make demand creation a waste. It's how you reach people before they start searching, and it tends to make the searching go your way later. It just comes second when money is tight.

3. Test small, then scale what works

• Pick two or three channels, not six.
• Give each a clear goal and a test budget you can afford to lose.
• Decide in advance how long the test runs and what result would count as success.
• Run each test long enough to learn something, usually several weeks.
• Move money towards the channel with the best cost per customer, and stop the one that isn't working, without sentiment.

4. Keep a steady slice for the long game

SEO and content take months to pay off, but they compound, and they're the channels least affected by rising ad prices. Once the short-term channels are working, set aside a steady portion for work that builds value over time. Consistency matters more than size here: a small amount every month beats a large burst followed by nothing.

5. Judge everything on cost per customer

Cheap clicks mean nothing if they don't turn into customers. Track what each channel costs per enquiry and per sale, and compare it with the figure you worked out above. That comparison, not habit, should decide next quarter's split.

When money is very tight

If the honest budget is close to nothing, the order above still holds, but the emphasis moves from spending money to spending time well:

• Fix the free foundations first. A clear homepage, a complete Google Business Profile, accurate listings and a habit of asking every customer for a review cost time, not money, and they make every later pound work harder.
• Ask your existing customers. Repeat business and referrals are often the cheapest customers to win, and a short, personal note to past customers can do more than a month of posting.
• Pick one channel and do it properly rather than three badly. For most local service businesses, that means being findable on Google; for businesses that sell something visual, it may be one social channel done consistently.
• Spend the first real money where it's easiest to measure. A small, carefully tracked search campaign tells you more, faster, than a scattering of boosted posts.

Revisit it every quarter

A budget set once a year and never looked at again is a guess with a spreadsheet attached. It works best as one box on a short plan you review every month, and our one-page marketing plan gives you a template for that. Every three months, ask it these questions:

• Which channel produced customers at the lowest cost?
• Which one produced none, and should it stop or change?
• Has the value of a customer, or the conversion from enquiry to customer, changed?
• Is there a foundation, such as the website or tracking, that is now holding everything else back?

Then move the money. Small, regular shifts based on evidence are how a modest budget ends up outperforming a larger one spent on habit.

If you'd rather not do the sums alone, our 90-day marketing plan sets the numbers and the order for your business, with a channel-by-channel split, for a fixed £2,500.