Cost per lead for trades: why every benchmark you've seen is wrong

Nobody agrees what a lead is

The first problem is definitional. To a directory, a lead is a homeowner's details released to you. To a Meta campaign, it is a completed instant form. To a Google Ads account, it might be a phone call lasting more than thirty seconds, or a form submission, or both counted once each. To a trade, it is someone who wants a quote for a job they do, in an area they cover, within a timescale they can meet.

Those are four different things with four different prices. A benchmark that averages across them is averaging apples, pears and the box they came in.

The seven ways a benchmark misleads

Caveat	Why it matters	Effect on the figure
Lead definition varies	Form fills, calls and released contacts are counted together	Meaningless average
Who published it	Lead sellers and ad agencies have reasons to show low or high numbers	Bias in whichever direction suits
Geography	London and the South East cost more than the North East for the same click	Regional spread hidden
Trade and job value	A £90 callout and a £15,000 extension are both 'a lead'	Cheap and dear trades averaged
Time of year	Boiler leads in January versus July	Seasonal swing flattened
Quality	Price shoppers, wrong areas, DIYers asking for advice	Cost per usable lead can be several times the headline
Response time	Leads lost to slow callbacks look like bad leads	Advertiser's own behaviour blamed on the channel

The regional problem in one example

Here is a hypothetical to show the scale. Two electricians run identical Google Ads campaigns for "electrician near me". One is in Croydon, the other in Hartlepool. The Croydon click costs more because more firms are bidding, and the Croydon enquiry converts differently because the homeowner has more alternatives. Their costs per lead could differ by a multiple, not a percentage, with the same ad, the same budget and the same skill.

A national benchmark sits somewhere between them and describes neither. If the Hartlepool electrician reads it and concludes his campaign is doing brilliantly, he stops optimising. If the Croydon one reads it and concludes his is a disaster, he switches it off. Both decisions are driven by a number that was never about them.

Cost per lead is the wrong numerator anyway

Even a perfect cost per lead would answer the wrong question. A trade does not want leads. It wants jobs, at a margin, from customers who pay. A channel producing cheap leads that never convert is dearer than one producing expensive leads that do.

The chain runs: spend, enquiry, quote, job, invoice value, margin. Cost per lead stops at the second link. The useful numbers are further down: cost per quote sent, cost per job won, and the ratio of marketing spend to invoiced revenue. Our explainer on how lead directory pricing actually works shows how to work cost per job out across platforms with different charging models.

The response time confound

The caveat most often ignored is the trade's own behaviour. Two firms buy the same leads from the same source. One rings back within five minutes; the other within a day. The first reports good leads, the second reports rubbish. The leads were identical.

This means any benchmark, and any comparison between your figures and a mate's, has to account for how quickly enquiries are handled. The arithmetic of what slow handling costs is worked through in the real cost of a missed call for a one-van trade. Until that is fixed, cost per lead measures your phone habits as much as your marketing.

What to track instead

Three numbers, kept for every channel you use, for at least three months before drawing conclusions.

• Cost per qualified enquiry: total spent on the channel divided by enquiries that were for a job you do, in an area you cover, from a real person. Not form fills. Not calls.
• Cost per job won: total spent divided by jobs actually booked and started. This is the figure to compare channels on.
• Marketing spend as a share of invoiced revenue by channel: what you spent on a channel divided by what the jobs it produced were worth. This shows whether a dear channel is bringing dear jobs.

When a benchmark is useful

Keeping those three numbers means marking every enquiry with its source and its outcome. A spreadsheet is fine. A CRM is better if you will use it. The discipline is the hard part, not the tool.

A published figure is a reasonable sanity check in one situation: when your own number is wildly different from every figure you can find, in either direction. Ten times cheaper usually means something is miscounted. Ten times dearer usually means something is broken, often the landing page or the targeting rather than the channel.

Beyond that, treat every benchmark as a description of someone else's business. The only cost per lead that matters is yours, measured properly, and compared with your own figure from last quarter. That comparison is also the basis for deciding how to pay anyone you bring in to help, which we cover in pay per lead, retainer or fixed fee: how trades pay for marketing.

Frequently asked questions