01The wrong question
"How much should we spend on PPC?" is usually asked as though there were a market rate: some going figure that businesses of your size pay. There is not. Budgets picked that way are round numbers chosen for comfort. A few hundred pounds because it feels cautious, a few thousand because it feels serious, and neither connected to anything.
Round-number budgets fail in one of two ways. Too small, and the account never produces enough data to learn from; six months later the conclusion is that ads do not work for you, when the honest conclusion is that you never bought enough evidence to know. Big enough but arbitrary, and there is no way to tell whether the results are good, because good was never defined.
The useful question runs the other way: what is a customer worth to us, and what can we afford to pay to win one? Answer that and the budget stops being an opinion. It becomes the output of arithmetic.
02Start with what a customer is worth
Everything hangs off one number: the value of a new customer. For a service business that means your average job or deal value, and the gross margin on it. If customers routinely come back, first-year value is a fairer measure than first-job value, though it pays to be conservative rather than flattering.
If you do not know these numbers, finding them is the first job, and it is a business job rather than an advertising one. Google Ads amplifies your economics. It cannot repair them, and no budget setting compensates for economics you have never measured.
03From customer value to affordable lead cost
A clearly hypothetical example, start to finish. Suppose you run a flooring company. Your average job is worth £2,400 and carries a 40% gross margin, so each job produces £960 of gross profit. You decide you are willing to trade a third of that profit to win a new customer, which sets a ceiling of £320 to acquire one.
Not every enquiry becomes a job. Suppose you win one in four. Then the most you can pay per enquiry is £320 divided by four: £80 per lead.
That is the whole method in one line: customer value, times margin, times the share of profit you will spend on acquisition, times your close rate. Your numbers will differ from these. The arithmetic will not.
04From lead cost to a monthly budget
Two more inputs turn an affordable lead cost into a budget: how many new customers you want (and can actually handle), and enough volume for the account to learn.
Continuing the example: you want eight new jobs a month. At one win in four, that needs thirty-two enquiries. Thirty-two enquiries at £80 each is £2,560 a month. That is the budget the goal implies. If it is more than you can commit, shrink the goal to match the money. Do not keep the goal and shrink the budget, because then the plan fails by design.
The same arithmetic reaches down to the click. Suppose your landing page turns one visitor in ten into an enquiry. Then you can pay up to £8 per click and still stay inside £80 per lead. If real clicks in your market cost more than that, the answer is rarely a bigger budget. It is a page that converts more of the clicks you already buy, which is the lever our landing page work exists to pull, or a sharper offer, or a better close rate. Improve any of them and every number upstream moves in your favour.
05The too-small-to-measure trap
The most common budgeting failure is not overspending. It is spending an amount that can never teach you anything: a budget that buys one or two enquiries a month.
At that volume you cannot separate skill from luck. A good month proves nothing, a bad month proves nothing, and every change is judged against noise. Automated bidding makes it worse, because bid strategies learn from conversion data; starve them and they never settle. The account is not failing. It is unmeasurable, which looks identical from the outside.
If the arithmetic from your own numbers implies a budget you cannot fund yet, do not spread a small budget thinly and hope. The better options, in order:
- Concentrate. One campaign, one service, one area. A small budget focused on your highest-intent searches beats the same budget spread across everything you do.
- Fix conversion rate first. Every improvement to the page raises what you can afford per click, before you spend a pound more.
- Wait. Not running ads is a legitimate decision. Running unmeasurable ads is not.
06Seasonality changes the answer
A single fixed monthly budget assumes your demand is flat across the year. For most businesses it is not, and pretending otherwise produces two predictable mistakes.
The first is judging a trough by the standards of a peak: the budget was set in a busy month, enquiries thin out in the quiet season, and the verdict is that the ads stopped working. The second is the mirror image: holding the cap steady through your peak, when demand is at its richest and the customers you are declining to fund are the easiest of the year to win.
Set the budget as an annual plan with a shape, not a flat line. Spend into the months that carry your year, ease back in the quiet ones, and when seasonality is strong, judge performance year on year rather than month against the month before.
07When to scale, and when not to
Scale when two things are true at once. First, you have a sustained run of stable results, a month or more, at a cost per customer you have explicitly decided is acceptable. Second, there is headroom: Google reports how often your ads missed auctions purely because of budget, and if your best campaign is losing impression share to budget, more money buys more of something already proven.
Then scale the boring way. Raise budgets in steps rather than leaps, give the bid strategy a week or two to resettle after each move, and watch cost per customer as you climb. Expect the curve to bend: each additional pound tends to buy slightly less than the one before, because you are reaching deeper into the auction. Scaling is finished when the marginal customer stops being profitable, not when the budget reaches a number that feels big.
And never scale to rescue underperformance. More spend multiplies whatever is already there. If the account is not working at £1,500 a month, £3,000 a month buys the same problem at twice the size.
That is the whole method: value, margin, close rate, lead cost, volume, budget, in that order. If you would rather have it run against your real numbers, it is the first thing we do in any Google Ads engagement, and a discovery call is where it starts.