Before anything else
If water is coming through a ceiling or a light fitting right now, stop reading this page. Turn the stopcock off, turn the electrics off at the consumer unit if the water is anywhere near them, and call whoever can get to you fastest. That may well be a sponsored result, and that is the correct decision. This page is for the hour after that, or for the leak you have been watching for a fortnight and have not yet done anything about.
What a click actually costs in London
Google Ads is an auction, and advertisers are charged per click. Google's own documentation is explicit about it: the actual cost-per-click is "the final amount you're charged for a click", set by the auction rather than by the advertiser. Nobody is charged for a job, a call or a customer. They are charged when a finger comes down on the link.
Here is what that costs on London leak and plumbing terms. These are Google Ads cost-per-click figures for London, pulled in September 2026. They are cost per click, not cost per lead and not cost per job.
| Search term | Cost per click, London, Sept 2026 |
|---|---|
| emergency leak repair | £103.54 |
| emergency plumber near me | £68.22 |
| leak repair | £48.34 |
| emergency plumber london | £46.47 |
| leak detection london | £45.60 |
| 24 hour plumber london | £40.78 |
| water leak detection london | £35.35 |
| trace and access london | £31.63 |
| ceiling leaking water | £21.07 |
| water leak repair | £18.52 |
| boiler leaking water | £16.88 |
| stopcock leaking | £15.07 |
The pattern in that column is worth sitting with. The cheap terms are the ones where you are still working out what is wrong. The expensive terms are the ones where you already know you need someone today. Urgency is what is being bid on, and you generate the urgency for free by having a leak.
The unweighted average of those twelve terms is £40.96. A real campaign is not spread evenly across them, so a well-run account that leans on the cheaper diagnostic terms will blend lower than that, and an account that only bids on the emergency terms will blend higher.
The arithmetic, and the assumption it turns on
A click is not a customer. To get from one to the other you need two numbers, and only one of them is measured.
Step one: clicks to enquiries
This is the number that does all the work, and it is an assumption, not a fact about any particular firm. Published benchmarks give a starting point. WordStream's Google Ads benchmark set, which is the most widely cited in the trade, puts the home and home improvement category's search conversion rate in the region of 7 to 10 per cent depending on which year's table you read. Unbounce, working from a different dataset of over 41,000 landing pages, reports a median conversion rate of 6.6 per cent across all industries for dedicated paid landing pages.
Both of those are averages across every kind of home-improvement advertiser, in every market, on every keyword. A broad emergency term in London is a harder case than the average, because a meaningful share of the clicks are people outside the service area, people who wanted a different trade, and people comparing before they decide. So the honest thing is to run the sum at more than one rate and let you pick.
There is a second reason to treat the published figures as an upper bound rather than a target. A "conversion" in an ads account is whatever the advertiser told the account to count. Commonly that is a form submitted, a phone call lasting more than half a minute, or a click on a call button, and each of those fires whether or not anyone ends up booking anything. A wrong number that stayed on the line for forty seconds is a conversion. So the benchmark rate measures clicks that became contacts, which is a lower bar than clicks that became enquiries a business could work with, which is in turn a lower bar than jobs. Every one of those gaps pushes the real cost per job upwards from the figures below, not downwards.
Step two: enquiries to booked jobs
The second assumption. Someone who rings about water coming through a ceiling is a serious enquiry, and a firm that answers the phone quickly will convert a high proportion of them. We have used 75 per cent below. That is generous to the advertiser, and it is deliberately generous, because a sum that only works when you stack the assumptions against someone is not worth publishing.
Putting the two together
Advertising cost per enquiry, at two different click costs:
| Assumed click-to-enquiry rate | At £46.47 a click | At £40.96 a click |
|---|---|---|
| 10% (near the top of the published benchmarks) | £465 | £410 |
| 6% (close to the all-industry median) | £775 | £683 |
| 3% (a hard, broad emergency term) | £1,549 | £1,365 |
Advertising cost per booked job, assuming 75 per cent of enquiries turn into work:
| Assumed click-to-enquiry rate | At £46.47 a click | At £40.96 a click |
|---|---|---|
| 10% | £620 | £546 |
| 6% | £1,033 | £910 |
| 3% | £2,065 | £1,820 |
If you think 75 per cent is too kind and prefer 50 per cent, every figure in the second table rises by half again. If you think the click-to-enquiry rate on a tightly targeted campaign is better than 10 per cent, the figures fall. The range across defensible assumptions runs from a few hundred pounds a job to a couple of thousand, and you can move within it by changing one number. That width is the honest answer, and anyone who gives you a single confident figure instead is guessing.
What does not move is the direction. At these click prices, advertising cost per acquired job on emergency leak terms in London is a material line in somebody's costs, and it is recovered from the customers who go ahead, because there is nobody else to recover it from.