Cost Per Lead vs Cost Per Customer
Most local service businesses quote a cost per lead that is far too low, because they are counting the ad spend and nothing else. The number that matters is not what a lead costs. It is what an acquired, profitable customer costs.
I am not going to give you a benchmark range. Acquisition cost varies so much between a roofing company and a dental practice — and between two roofing companies in the same city — that any "typical" figure is noise. What is portable is the method. Here it is.
The formula
Total acquisition cost for a channel, over a period:
``` media spend
- agency or internal labor
- software and platform costs
- creative and landing-page costs
- lead qualification costs
- sales follow-up costs
- cost of invalid or shared leads
= total acquisition cost ```
Most businesses stop at the first line. The rest is where the money actually goes.
The five numbers, not one
Divide that total by a different denominator each time and you get five very different figures. They answer five different questions.
| Metric | Denominator | What it tells you |
|---|---|---|
| Raw lead cost | Every enquiry | How efficiently the channel generates volume |
| Qualified lead cost | Enquiries that are real, in-area and in-service | Whether the volume is worth anything |
| Booked-appointment cost | Appointments actually on the calendar | Whether your follow-up works |
| Acquired-customer cost | Customers who paid | The number most owners think they already know |
| Gross-profit-adjusted cost | Customers, against gross profit not revenue | Whether the channel makes you money |
The gap between the first and the fourth is where most marketing decisions go wrong. A channel with a cheap raw lead cost and a terrible qualification rate loses to an expensive channel with good intent, every time.
A worked example
This is a hypothetical, not a benchmark and not client data. The numbers are chosen to be easy to follow. Substitute your own — the structure is the point.
A business spends $3,000 on media in a month. It also pays $680 in labor to run the campaigns and answer enquiries, $170 in software, and $350 amortised on creative and a landing page. Total: $4,200.
That produces 100 enquiries. Of those, 60 are real and in-service-area. 30 book an appointment. 10 become customers.
- Raw lead cost: $4,200 ÷ 100 = $42
- Qualified lead cost: $4,200 ÷ 60 = $70
- Booked-appointment cost: $4,200 ÷ 30 = $140
- Acquired-customer cost: $4,200 ÷ 10 = $420
If the average job bills $2,400 at a 40% gross margin, each customer contributes $960 of gross profit against $420 of acquisition cost. That channel works. If the margin were 15%, each customer contributes $360 against $420 — the channel is losing money while the raw lead cost still looks fine.
Shared lead platforms need one more line
A shared lead is sold to several businesses at once, so some proportion is unwinnable regardless of how well you sell. When you compare a shared platform against a channel you own, that has to be in the denominator: divide by the leads you could realistically win, not the leads you were billed for.
Do not compare a shared platform's per-lead price against your website's per-lead cost and conclude the website is expensive. They are not the same unit.
What this method will not tell you
- Attribution across touches. A customer who found you in an AI answer, read two pages, and called you three weeks later after seeing an ad will be credited to one channel by most tooling. This method measures channels; it does not resolve multi-touch attribution.
- Lag. Long-consideration trades — roofing, windows, legal — book customers months after the spend. A single month's figures will mislead you. Use a window that matches your actual sales cycle.
- Lifetime value. Recurring work such as pool service or bookkeeping justifies a much higher acquisition cost than a one-off job. If your work recurs, run the last row against lifetime gross profit rather than first-job gross profit.
What I do with this
When I take on a partnership, this is one of the first things I build, because almost nobody has it. Until the tracking exists you cannot tell whether a channel is working, which means every decision about where to put money is a guess. Getting conversion tracking honest usually comes before anything clever.
Written and reviewed by Vladimir Kamenev. Last reviewed 16 August 2026. The worked example above is hypothetical and is labelled as such; it is not client data and not an industry benchmark.
Cost per customer is the number I report on for every client, because it is the only fair way to compare small business lead generation channels.
Frequently Asked Questions
What is a good cost per lead for a local service business?
There isn't a portable number, and I'd distrust anyone who gives you one. Acquisition cost varies enormously between trades, between markets and between two businesses in the same trade and city. The useful question is whether your acquired-customer cost is comfortably below the gross profit that customer produces. Work that out for your own business rather than comparing against an industry average.
How do I account for indirect costs?
Add the labor spent running campaigns and answering enquiries, the software and platform fees, and the amortised cost of creative and landing pages to the media spend before you divide. Most businesses count only media spend, which is why their reported cost per lead is far lower than the real one.
Can I use cost per lead to compare channels?
Only if you compare the same unit. A shared lead platform sells the same enquiry to several businesses, so some of it is unwinnable no matter how well you sell; divide by the leads you could realistically win, not the leads you were billed for. Comparing a shared per-lead price against an owned channel raw lead cost will mislead you.
How often should I recalculate?
Monthly for fast-cycle work, but use a window that matches your actual sales cycle. For long-consideration trades like roofing, windows or legal work, customers book months after the spend, so a single month read in isolation will point you in the wrong direction.
Where does customer lifetime value come in?
If your work recurs — pool service, bookkeeping, maintenance contracts — run the final calculation against lifetime gross profit rather than first-job gross profit. Recurring work justifies a much higher acquisition cost than a one-off job, and treating them the same will cause you to switch off channels that are actually profitable.
What is the most common mistake?
Optimising the raw lead cost. A channel with cheap leads and a poor qualification rate loses to an expensive channel with good intent almost every time. Track the gap between enquiries and acquired customers before you touch the budget.