Almost every small business owner ends up asking the same question: how much should I spend on marketing? And almost always gets the same unhelpful answer: it depends. It's true that it depends, but that doesn't excuse you from deciding on a figure, because without one you either spend blindly or spend nothing.
This article won't give you a magic number. It gives you a way to reason it out, three methods to set a starting figure, a formula to work out how much you can pay for each customer, and a way to split the money between your website, SEO, advertising and content without spreading it thin. Every number you'll see is an example with round figures, meant to show the logic: they are neither real data nor market data.
Why there's no universal figure
Two businesses with the same turnover can need very different budgets. One starts with an outdated website, no Google profile and no idea where its customers come from; the other already has the basics and only needs to attract more. One sells a 20 € product that's bought once; another sells a 2,000 € service that creates a relationship lasting years. Each difference moves the figure.
That's why rules of the type invest such and such a percentage of your turnover are, at most, a starting point. What really determines how much it makes sense to invest is the economics of your business: how much you earn per customer, how much it costs to win one and how long you can wait for the investment to pay off. And if you can't answer them yet, part of the first budget should go on measuring.
Three methods to set the budget
They aren't mutually exclusive. The sensible approach is to work out all three and compare the results.
Method 1: a percentage as a starting point. It means setting aside a share of the year's turnover or profit. Illustrative example: a shop with a turnover of 200,000 € a year decides to start from 5%, a figure chosen only for the example. That gives 10,000 € a year, about 830 € a month. With profit it works the same way: if that business earns 40,000 €, 10% would be 4,000 €. It's quick, but it has a limit: it ignores your goal. Someone who wants to grow or is starting from zero usually needs more than the rule gives, and someone who already lives off word of mouth may need less.
Method 2: start from the goal and from what it costs to win a customer. Illustrative example: you want 30 new customers a year through digital channels and, after doing the sums (you'll see how in the next section), you conclude you can pay up to 200 € for each one. That's 30 × 200 = 6,000 € a year as a ceiling, that is, 500 € a month. It's more precise than the percentage, but it requires knowing your numbers, and it's worth leaving some slack: the first customers tend to cost more while everything is being tuned.
Method 3: cap it at what you can afford to lose during the trial. Every new channel has a period in which you spend before knowing whether it works. The question is: how much can I spend over six months without my cash flow suffering if nothing comes out of it? Illustrative example: 3,000 € over six months, or 500 € a month. It's your risk ceiling, not your ideal budget. And before you start it's worth writing down what result would make you continue, adjust or stop.
How to combine them: if in the example method 1 says 830 € a month, method 2 says 500 € and method 3 says 500 €, you start at 500 € and go up when the data justifies it.
How much you can pay per customer: CAC, LTV and margin
Three concepts explain almost everything:
- CAC (customer acquisition cost): what you spend on winning new customers divided by the number of new customers you won. It includes everything it costs to win them (ads, agency, tools, content), not just the money you put into advertising.
- LTV (customer lifetime value): what a customer brings you over their whole relationship with you, not just on the first purchase.
- Margin: the part of the price you keep after paying the direct costs of producing or delivering what you sell.
They combine in a simple formula:
Total margin per customer = average ticket × number of purchases × margin (%)
That result is the ceiling on what you can pay to win a customer without losing money. Illustrative example: your average ticket is 80 €, a customer buys three times over their relationship with you and your margin is 40%. Then 80 × 3 × 0.40 = 96 €. If bringing in a customer costs you less than 96 €, you make money on them; if it costs more, you lose it. And 96 € is the limit, not the target: you have fixed costs and you want profit, so the prudent thing is to aim well below, for example half (48 €).
Two precautions. First: LTV is an estimate; if the business is new, use cautious assumptions and correct them with real data. Second: repeat purchases arrive over time, so don't count on them to pay for ads today if your cash flow can't handle the wait. And notice: if loyalty work gets the customer to buy four times instead of three, the ceiling rises to 80 × 4 × 0.40 = 128 €. Retention is also money well spent.
How to split it by layers and by phases
Splitting well doesn't mean dividing the money into equal parts, but putting it where it's needed at each moment. A practical way to see it is three layers:
- Foundations: a website that converts (clear, fast, good on mobile and with a visible call to action), a complete and well-kept Google profile, and analytics to know what's happening. Without this, every euro you spend afterwards works less.
- Acquisition: local and content SEO, which is medium term, and paid advertising, which gives quick results for as long as you keep it running.
- Retention: email, a simple CRM so you don't lose track of your customers, reviews and repurchase strategies.
The weight of each layer changes over time. The following table is an illustrative example based on any monthly budget, not a recommendation or market data:
| Layer | Phase 1 (months 1–3) | Phase 2 (months 4–9) | Phase 3 (from month 10) |
|---|---|---|---|
| Foundations (website, Google profile, analytics) | 50% | 15% | 10% |
| Local SEO and content | 15% | 35% | 35% |
| Paid advertising | 25% | 30% | 30% |
| Retention (email, CRM, reviews) | 10% | 20% | 25% |
Look at the logic rather than the percentages. Foundations weigh heavily at the start because a large part is one-off work that afterwards only needs maintenance. SEO and content gain weight once the base is ready, because they take the longest to pay off. Advertising comes in early and in moderation, because it gives quick data, but it only makes sense once the website is ready to receive visits.
What accumulates, how long to wait and which mistakes to avoid
There's an important difference between two kinds of spending. The website, SEO, content and your customer database are an investment that accumulates: they keep working for you even if you stop investing for a while. Ads are spending that disappears when you stop: as soon as you pause the campaign, the visits that depended on it end. That doesn't make them bad: they're a quick, measurable lever. But if your whole budget is advertising, you're essentially renting your customer base. The part that stays in your business should never be zero.
How long should you give each channel before deciding? There are no guaranteed timeframes, because they depend on your sector, your competition and your starting point. There are, however, reasonable criteria:
- Paid advertising: the first signals (clicks, enquiries) arrive early, but it's worth allowing at least a couple of adjustment cycles before judging, and judging by cost per enquiry or per customer, not by clicks.
- Local SEO and content: it's a bet of several months at the very least. In the meantime look at intermediate indicators: visits from Google, rankings, calls from your profile.
- Email and reviews: the effect is gradual and shows up in repeat purchases and new ratings.
What matters is to set the review date and the criteria before you start, not when impatience kicks in. And these are the mistakes that repeat most often:
- Spreading too little money across too many channels. With 500 € a month split across five channels you get useful data from none of them. Better two channels well fed.
- Pausing just before it works. Without a prior criterion, it's easy to quit just when it was starting to pay off.
- Measuring only clicks. Clicks don't pay invoices. Measure calls, forms, quotes requested and sales.
- Paying for visibility without a website that's ready. Sending visits to a slow, confusing page or one with no clear way to get in touch is throwing money away.
- Thinking the budget is only advertising. The website, content, tools and retention are also marketing and also cost money.
What else to include in the budget and how to review it every quarter
Many budgets fall short because they only consider the agency's invoice or the money for ads. Add up these as well:
- Tools: subscriptions for email, analytics, appointment booking, CRM.
- Content: texts, videos, graphics.
- Photos: good images of your premises, your team and your products.
- Internal time: yours or your team's, to answer messages, approve pieces and provide material. It's a real cost even if it appears on no invoice.
- Taxes and platform commissions: be clear on whether your figures include VAT (whether you can deduct it depends on your situation, so check with your accountant) and count the commissions of payment gateways, marketplaces or booking platforms.
Also leave a small cushion for the unexpected.
Once things are running, review the split every quarter with a few indicators, always the same ones and recorded in the same place:
- Enquiries or orders received, by channel.
- CAC by channel, compared with your ceiling of margin per customer.
- Website conversion: what proportion of visits ends in contact or a purchase.
- Customers who come back and new reviews.
- Real margin on what was sold.
The underlying question is simple: which channels bring in customers below your limit and which don't? Move part of the money from what doesn't work to what does, in small steps. Avoid abrupt turns based on a single month, which may be atypical.
How to talk to an agency or a freelancer without ending up with a closed package
A good conversation starts by giving them context, not just a figure: your goal, your approximate margin, your risk limit and the trial period you have in mind. With that they can propose something tailored to your budget instead of fitting you into a standard package. Some questions that help:
- What would you do with half of this budget and what would you leave out?
- What will be measured, how often is it reviewed and by what criteria do we decide to continue or change?
- Whose name will the ad accounts, the analytics and the Google profile be under? They should be yours.
- If I stop paying in a year, what do I keep and what disappears?
Be wary of packages that don't ask about your business and of any promise of guaranteed results: nobody can honestly assure them.
Finally, public grants and subsidies for digitalisation can be a useful complement, but not a foundation on which to build the budget. Calls change, so always check the one currently open with the relevant official body or with your adviser, and don't spend counting on a grant you haven't been awarded yet.
Frequently asked questions
How much should I invest if I've just opened my business?
There's no single figure. At the start, an important part usually goes on the foundations (website, Google profile, analytics) and on measuring. Work out the three methods in this article, start with an amount you can lose without putting your cash flow at risk and increase it when the data supports it. Results and specific timeframes can't be promised.
Is it better to invest in SEO or in advertising?
They don't compete with each other. Advertising gives data and visits quickly, but it stops when you stop paying; SEO takes longer, but it accumulates. The usual approach is to combine both, with a website ready to turn visits into customers.
How often should I change the split?
A good rhythm is to review it every quarter with the same indicators and make small changes. Deciding on the data of a single week, or of an atypical month, usually leads to changes that later have to be undone.