Most budget advice starts in the wrong place: what you can afford, or what a benchmark says an industry spends. Neither tells you whether the spend will make money, because both are somebody else's numbers.
The useful method runs backwards from what a customer is actually worth to you.
We publish no cost benchmarks here on purpose. Click costs vary by market, geography, seasonality and competitor behaviour, and a figure quoted in an article is out of date before it is read. Every number below comes from your own business.
Working backwards
Five steps. You need real figures for the first three; if you do not have them, that is the finding.
1. What is a customer worth? Gross margin on a typical customer, over the realistic relationship — not headline revenue. If most customers buy repeatedly, use the margin across that lifetime.
2. What can you spend to acquire one? A share of that margin. The rest is your profit and overhead. This is your allowable cost per acquisition, and it is a decision, not a calculation.
3. What is your close rate on enquiries? Of the enquiries reaching you from a comparable channel, what proportion become customers? If you have never measured it, measure it before spending — this is the number that converts advertising cost into business outcome.
4. Derive the allowable cost per enquiry.
allowable cost per enquiry = allowable cost per acquisition × close rate
If you can spend a given amount to win a customer and one enquiry in five becomes one, you can spend a fifth of that per enquiry.
5. Multiply by the enquiries you need.
monthly budget = allowable cost per enquiry × enquiries needed per month
That is a budget derived from your economics. It may be uncomfortably large or surprisingly small — either result is informative, and both beat a number chosen because it felt reasonable.
Getting step 3 right requires knowing which enquiries actually became customers, which is why measurement precedes spend in
paid campaigns built from unit economics rather than benchmarks.The viable minimum
There is a floor, and it is not a currency figure. It is the point at which the campaign produces enough conversion data to be judged.
Two things depend on that data:
- Your ability to decide. A handful of conversions cannot distinguish a working campaign from a lucky fortnight
- Automated bidding. Smart Bidding strategies need consistent conversion volume to optimise rather than react to noise. Below that, they amplify randomness
So the honest minimum is: enough spend that within a reasonable window you accumulate enough conversions to draw a conclusion. Where clicks are expensive and close rates low, that floor is high. Where a conversion is cheap and the allowable cost per acquisition generous, it is much lower.
If your budget cannot reach that floor, paid search may be the wrong channel for now — and that is a legitimate finding rather than a failure. Spending below the threshold buys activity, not information.
Phasing
| Phase | Purpose | Judge on |
|---|---|---|
| Validation | Does this convert at all, at any cost? | Conversions existing, search terms being relevant |
| Efficiency | Can it convert inside the allowable cost? | Cost per conversion against your number |
| Scale | How much profitable volume exists? | Cost per conversion holding as spend rises |
Most accounts skip validation and start at efficiency, judging a two-week-old campaign on cost per acquisition before it has established whether the offer converts from paid search at all. Those are separate questions and they need separate windows.
Do not judge during a learning period. Bid strategy changes need time before the resulting data means anything, and evaluating mid-change restarts the process — so you never observe a stable result.
When to increase
One clear signal: a campaign consistently limited by budget while converting inside your allowable cost. That is demand you are declining to serve.
Increase gradually. Sharp budget changes destabilise bidding systems, and the resulting turbulence gets misread as the increase failing.
Expect cost per conversion to rise as you scale. The cheapest, most relevant impressions are captured first; additional spend reaches progressively weaker matches. Decide in advance how much rise is acceptable — the point at which cost per acquisition exceeds your allowable figure is the ceiling, and it is a real ceiling rather than a target to push through.
When to stop
Decide the stop condition before you start, while you are still objective about it.
Reasonable conditions:
- Cost per acquisition exceeds allowable by a defined margin, sustained over a full measurement window
- Sufficient spend has produced insufficient conversions to conclude anything
- Search terms show the intent is not there, whatever the settings
- The constraint turns out to be elsewhere — the offer, the site, response times
That last one matters more than the budget question. If enquiries arrive and are handled slowly, or the page converts poorly, more budget multiplies an existing leak. Fixing the leak is cheaper than funding it.
Three ways budgets get wasted
Spread too thin. A modest budget across five campaigns produces five campaigns with insufficient data. Consolidate until each has enough volume to be judged, then separate.
Running when nobody answers. For any business converting by phone, ads running outside answering hours buy clicks that cannot convert. This is checkable in an afternoon.
Buying terms that could never convert. The search terms report shows what you actually paid for, and it routinely contains job seekers, students and researchers. The Google Ads audit checklist covers this at stage three, after tracking and structure.
Before you set any number
Two prerequisites, both non-financial:
- Conversion tracking verified end to end. A budget decision made from wrong conversion data is worse than no decision — see conversion tracking that actually works
- A known close rate, or a plan to establish one. Without it, step 3 is a guess and the whole calculation inherits it
If neither exists yet, that is where the first two weeks go. It is unglamorous and it is the difference between a budget and a number.
Want this handled properly?
You now know what the work involves. If you would rather not do it yourself, that is what we do.