The short version
- The right budget is not a percentage of turnover. It is whatever buys enough clicks to produce a statistically useful number of enquiries each month.
- Work backwards from your closing rate and average order value, not forwards from what you can spare.
- Underfunding a campaign is worse than not running one. You pay for the learning period and stop before it pays you back.
- Management fees and ad spend are separate budgets. Treat them separately.
Nobody asks this question in the abstract. They ask it because they have a number in mind and want to know whether it is sensible. So let us deal with that directly: the question is not how much you should spend, it is whether the amount you have in mind can produce a result you could learn anything from.
Work backwards from the sale
You need four numbers. Most businesses can produce three of them immediately and have to estimate the fourth.
- Average order value, or better, average gross margin per customer. Margin is the honest number. Revenue flatters everything.
- Your closing rate on enquiries. Out of ten enquiries, how many buy?
- The most you would happily pay for one customer. Not the most you could survive paying. The number that leaves the deal clearly worth doing.
- Your landing page conversion rate. If you do not know, assume something modest and correct it with real data after a month.
Here is the arithmetic, using illustrative figures rather than any claim about your business. Say your average customer produces £4,000 of gross margin and you close one enquiry in five. One customer therefore needs five enquiries. If you are content to spend up to a quarter of the margin acquiring the customer, that is £1,000 per customer, or £200 per enquiry.
Now the other direction. If your landing page converts one visitor in twenty into an enquiry, £200 per enquiry means you can afford roughly £10 per click. If clicks in your sector cost £4, you are comfortable. If they cost £14, the maths does not work at your current conversion rate and closing rate, and no amount of clever campaign management fixes that. You would need to improve the page, the offer, or the qualification of who you attract.
Running the numbers above tells you whether paid search is viable for you at all, before you spend anything. It is a twenty minute exercise and it has talked us out of taking on work more than once, which is a better outcome for everybody than six months of disappointment.
The minimum useful budget
Separately from what you can afford, there is a floor below which a campaign cannot teach you anything.
Automated bidding needs a reasonable flow of conversions before it behaves sensibly. Below that, you are running a campaign that is permanently in its learning phase, and the numbers swing wildly month to month for reasons that have nothing to do with your marketing. You end up making decisions on noise.
As a working rule, a starting budget needs to buy enough clicks per month to produce a meaningful number of enquiries, not two or three. What that costs depends entirely on click prices in your sector, which is why any agency quoting you a universal minimum is guessing. In competitive B2B and construction categories, click prices are high enough that a small budget produces a handful of clicks a day, and a handful of clicks a day will not tell you whether your campaign works.
If the honest minimum is more than you want to commit, the sensible move is usually to narrow the campaign rather than thin the budget. One service, one location, properly funded, beats five services everywhere on the same money.
Budget for testing, separately
A campaign that never changes stops improving. Testing is not an optional extra, but it does have a cost, and it is worth naming it.
Every variation you want to test needs enough traffic behind it to produce a result you can trust. Splitting a small budget across four landing page variations gives you four sets of numbers that all look different and none of which mean anything. It is better to test one thing at a time, properly funded, and accept that this takes months rather than weeks.
Plan on funding each variation with enough spend to reach a sensible number of conversions before you judge it. If that is not affordable, test fewer things.
Want the maths run on your numbers? Send us your average order value, closing rate and the sector you are in. We will tell you whether paid search stacks up for you and roughly what it would take to find out properly.
Management fees are a separate line
A common and expensive confusion is to treat one pot of money as covering both the ads and the people running them. It creates a perverse incentive: the more carefully somebody manages your account, the less there is left to spend on it.
Keep them separate on the budget sheet. Ad spend goes to the platform. Management is a service you buy, priced against the work involved rather than as a slice of your media. Our own pricing is set out on that basis, and our Google Ads management page explains what the work includes.
A useful sense check: if management costs more than the spend it manages, the account is probably too small for an agency and you would be better served by a fixed-scope setup and a quarterly review.
What to do in the first ninety days
Budget planning is easier if you know what the money is buying at each stage.
- Weeks one to two. Tracking first. Conversion actions, call tracking, CRM connection. Do not turn on spend before you can measure it. Everything after this depends on it.
- Weeks three to six. Launch narrow. Your clearest service, your best geography, exact and phrase match. Expect the cost per enquiry to be high and unstable. This is the learning period and it is not wasted.
- Weeks seven to twelve. Prune and widen. Negative keywords from real search terms, budget shifted towards what produces qualified enquiries rather than cheap ones. This is where the account starts to look like an asset.
Judging the campaign at week four is the most common and most costly mistake. Almost every account looks disappointing at week four.
Common questions
Is there a percentage of revenue I should be spending?
Rules of thumb like that are built for planning at board level, not for setting a campaign budget. They tell you what is affordable, not what is effective. Use the backwards calculation instead.
Should I start small and scale up if it works?
Starting narrow is sensible. Starting thin is not. Reduce the number of things you are advertising rather than the money behind each one, so the campaign you do run has enough data to be judged fairly.
What if my competitors have far bigger budgets?
Then compete where they are careless. Long-tail search terms, specific services, specific locations, and a landing page that speaks to a narrow audience will beat a generic campaign with more money behind it. This is also where organic search earns its keep, because the ground you win there does not have to be rebought every month.
Does the same maths apply to paid social?
The structure does, but the numbers move. Social clicks are usually cheaper and colder, so expect a lower click cost and a lower conversion rate, and check the whole chain rather than assuming the cheaper click is the better deal.
