Ask most agencies how much you should spend on advertising and you get “it depends”. It is technically true and completely useless. Yes, it depends. On what, and how do you actually work it out? That is the question you were asking, and you deserve a real answer.
The good news is there is a proper method. You can work backwards from a few numbers you already half-know and arrive at a budget that is grounded in your own business rather than plucked from the air. Let me walk through it.
Work backwards from what a customer is worth
Start at the end, with the value of a customer, not the cost of a click.
Take the average amount a customer is worth to you. If you are a one-off purchase business, that is the profit on a typical sale. If people come back, use the profit over the life of the relationship, which is usually much higher and is the number that really matters. A boiler service that costs 90 pounds looks small until you notice the same customer returns every year for a decade.
Once you know what a customer is worth, you can decide what you are willing to pay to win one. That figure is your target cost per acquisition. If a customer brings you 400 pounds of profit over time, you might happily spend 80 to acquire one. The ceiling is clear: you cannot pay more to win a customer than they are worth, and you want a healthy margin below it.
Turn that into a minimum viable budget
Now bring in conversion rates, and be realistic rather than hopeful.
Suppose a well-built landing page turns 1 in 20 clicks into an enquiry, and you close 1 in 3 of those enquiries into paying customers. That means it takes roughly 60 clicks to get one customer. If clicks in your industry cost around a pound, one customer costs about 60 pounds in ad spend. Sits comfortably under an 80 pound target, so the maths works.
Then ask how many customers you actually want. If you want ten new customers a month at 60 pounds each, your minimum viable budget is around 600 pounds a month, plus management. That is the number. Not “it depends”. A figure you can defend, built from your own value and honest conversion rates.
Run it for your business and the answer might be higher or lower, but the shape is the same:
- Customer value sets the ceiling on cost per acquisition.
- Cost per click and conversion rates set what a customer actually costs.
- The number of customers you want sets the total budget.
Why five pounds a day on Google search is usually wasted
This method explains one of the most common mistakes we see. A business puts five pounds a day, roughly 150 a month, into Google search and concludes after a month that “Google Ads doesn’t work”.
Look at the maths. At a pound a click, 150 pounds buys 150 clicks in a month. If it takes 60 clicks to make a customer, that is around two or three customers, spread thinly, on some days none at all. There is not enough volume to draw any conclusion, and worse, the account never gathers enough data to improve.
Respect the learning phase
Modern ad platforms lean heavily on automated bidding, and automation needs data to work. When a campaign is new, it goes through a learning phase where the system is still figuring out who to show your ads to and what to pay. During that stretch, results are unstable and usually worse than what follows.
Rough guide: a campaign wants a decent run of conversions in a couple of weeks to get through learning and settle down. If your budget only produces two or three conversions a month, you never escape learning, and the platform never gets good at your account. You are permanently stuck in the worst phase of the campaign. Underfunding does not give you a cheaper version of advertising. It gives you a version that cannot work.
Underfunded or genuinely wrong
The final judgement is knowing when a channel truly does not suit you, versus when you simply never gave it enough to work with.
Ask yourself honestly:
- Did I fund enough clicks to produce a meaningful number of customers, not just a handful?
- Did I run it long enough to clear the learning phase and then gather real data?
- Was the tracking correct, so I actually know what it produced?
- Was the landing page and offer good enough to convert the clicks I paid for?
If the answer to all four is yes and the numbers still do not add up, fair enough. That channel is genuinely not for you, and you can walk away with confidence. But if any answer is no, you have not tested the channel. You have tested a starved, broken version of it and blamed the channel for the result.
Getting this right at the start saves a lot of wasted money and a lot of wrong conclusions. If you want help sizing a realistic budget before you commit, that is one of the first things we work out with a client, and we would rather tell you a channel needs more than you want to spend than let it fail quietly.