Work backwards from the revenue you want

Start with the outcome. Suppose you want an extra £120,000 in revenue over the year. If an average customer is worth £4,000 and you win around one in four qualified leads, you need about thirty new customers and around 120 qualified leads. These figures are illustrative, but the method works for any business: goal, average value, close rate, leads needed.

Once you know how many leads you need, you can ask what each can cost before the work stops being profitable. That is your ceiling, and it makes every channel easier to judge.

Know your margins before you set a limit

A budget based on revenue ignores what each sale leaves behind. Use gross margin: if a customer is worth £4,000 in revenue but only £1,600 in gross profit, that is closer to what you can afford to spend winning them. Over time, also look at repeat business. A customer who comes back several times can justify a higher acquisition cost than a one-off buyer.

Split one-off setup from ongoing spend

Some costs happen once or occasionally: a new website, brand work, tracking, photography. Others repeat every month, such as advertising, content, email tools and agency fees. Keeping the two apart stops a big launch cost distorting your monthly figures, and helps you compare suppliers properly. Ongoing costs usually include:

• Media spend on search, social or other advertising

• Agency or freelancer fees

• Software for email, analytics, scheduling and reporting

• Content production, such as writing, design and video

• Your own team's time, which is easy to overlook

Match spend to the stage of your funnel

Money spent bringing in traffic is wasted if the website cannot convert it. Before you raise ad spend, check that visitors have a clear route to enquire, and that follow-up is fast. Our guide to building a marketing funnel for a service business shows how to find the weak step. Fixing a leak is often cheaper than buying more visitors.

Choose channels by speed and sales cycle

Paid search, run through Google Ads, can produce enquiries quickly, but you pay for each visit. Organic search takes longer and results are not guaranteed, though it can keep working without a payment per click. Our comparison of SEO versus PPC explains how to choose between them. If your customers take months to decide, favour channels that build trust over time rather than only chasing instant clicks.

Leave room to test

Not every channel suits every business. Set aside roughly a tenth to a fifth of the budget for experiments, as a rule of thumb rather than a law. Give each test a clear question, a fixed budget and a date to review it. A small test is cheaper than committing a whole year to something unproven. Make sure conversions are tracked in Google Analytics so that you can tell what worked.

Review monthly and adjust quarterly

A budget is a hypothesis. Each month, compare spend with qualified enquiries and cost per lead. Each quarter, move money from what is underperforming to what is working. Our guide to measuring marketing ROI without misleading yourself covers the pitfalls to watch for, especially when platforms report their own results. Peer groups such as the Federation of Small Businesses can also help you sanity-check what similar firms are spending.