How much should you spend on Google Ads? Work it out from the click, not from what is spare

Spend enough to buy the clicks it takes to learn, and no more than the arithmetic supports: the cost of a click in your sector, times the clicks it takes to get an enquiry, has to sit below what an enquiry is worth to you. For most UK small businesses that puts the floor at a few hundred pounds a month, because below it the account never collects enough data to improve, and the ceiling is set by cost per lead rather than by ambition. The budget you have spare is not a number in the sum.

Start from the click, not from the budget

Every Google Ads budget conversation should start with one figure: what a click costs for the searches you want. It varies more by sector than by anything else. WordStream's 2025 benchmark report, which pools thousands of mostly American accounts, put the average Search cost per click across all industries at a little over five dollars, with legal services, home services and finance well above that and retail, travel and hobbies well below. UK costs in the same sectors tend to land in a similar range in pounds, which is the only reason to quote a dollar figure at all; the number that matters is the one for your searches in your area.

You can see that number without spending anything. Google's Keyword Planner, inside any Google Ads account, shows a top-of-page bid range for each search in a chosen location. Put in the five searches a customer would type, set the location to the area you serve, and read the range. It is an estimate, and real CPCs land above or below it depending on Quality Score, but it is your sector and your area rather than an average of everyone else's.

The sum, step by step

1. Decide what a customer is worth to you. Not revenue; the margin on an average job or order, or the lifetime value if they come back. Call it £400.
2. Decide how many enquiries you win. If two in five become customers, an enquiry is worth 40% of £400, which is £160. That is the most the account can spend to produce one, including the management fee, before it loses money.
3. Estimate the share of clicks that become an enquiry. A Search campaign landing on a page built for the search converts somewhere between three and eight per cent for most service businesses; a homepage does worse. Start at five and correct it with real data after the first month.
4. Divide the cost per click by that rate. At £2 a click and 5%, an enquiry costs £40 in ad spend. Add the management fee spread over the month's enquiries and compare the total with the £160 an enquiry is worth.
5. Decide how many enquiries a month you want, and multiply. Twenty enquiries at £40 is £800 of ad spend. That is the budget, and it came from the click, not from the bank balance.

The floor: enough clicks to learn

Smart Bidding, which is what most accounts should be using once conversion tracking is proven, learns from conversions, and Google's own guidance is that a campaign needs roughly thirty of them in a month before a target cost per action behaves. At a 5% conversion rate that is six hundred clicks, and at £1.50 a click that is around nine hundred pounds. Below that the account still works, but it works on manual or Maximise Clicks bidding with tighter match types, and improvement comes from the person reading the Search Terms report rather than from the algorithm.

There is a smaller floor inside each ad group. Judging an ad group on twenty clicks is judging noise. Give each one at least a hundred clicks before deciding it does not work, which on a small budget means fewer ad groups, not more patience.

The ceiling: impression share and the marginal click

The account tells you when more money would buy more customers. The Search campaign reports the share of eligible searches your ads were shown for and why they were not: lost to budget, or lost to rank. Share lost to budget with a cost per enquiry under what an enquiry is worth is a signal to spend more, and the report should say so with the number. Share lost to rank means the extra pound buys nothing until Quality Score, the ads or the page improve.

Even when budget is the limit, the last pound buys a worse click than the first. Raising a budget also raises the bids the strategy is willing to pay, and the cost per enquiry drifts up as the campaign reaches for searches it was previously outbid on. Increase in steps of twenty to thirty per cent, read the cost per enquiry a fortnight later, and stop when it crosses the line the first step of the sum drew.

Where the number goes wrong

• The budget is whatever was left over, so it is either too small to learn from or too large for the searches that exist in the area.
• The budget is set by an agency paid a percentage of it, and the recommendation to increase arrives every quarter regardless of the impression share report.
• Google's suggested daily budget is accepted. It is calculated to spend, not to convert, and it is usually several times what the searches justify.
• The management fee is left out of the sum, so an account that looks profitable on ad spend alone is not.
• The same amount is spent every month of the year, so the quiet months are overpaid and the busy ones are capped by noon.

A test you can do tonight: open Keyword Planner, type five searches a customer would make, set your area, and write down the top-of-page bid range. Divide the top of the range by 0.05. If that number is bigger than what an enquiry is worth to you, Google Ads is not your first spend, and no budget changes that.

The estimates in the sum get replaced by real figures in the first month: the actual CPC, the actual conversion rate, the actual enquiries. The budget is then recalculated, not defended.