What a lead actually costs

Three numbers, one of which matters

Every advertising conversation involves three costs, and people slide between them without noticing. Cost per click is what you pay for a visit. Cost per lead is what you pay for an inquiry. Cost per customer is what you pay for someone who actually gave you money.

They can be wildly different from each other, and only the third one can answer the question you actually have, which is whether to keep spending. A campaign can have an excellent cost per click, a decent cost per lead, and still lose money on every customer it produces.

Cost per click is a vanity number

It is the number platforms show most prominently and the one agencies most like to report, because it is the easiest to improve and the least connected to your bank balance.

You can drive it down almost at will by bidding on cheaper, vaguer keywords. Those clicks cost less because they are worth less. A campaign whose click cost halved while its customer count stayed flat has not improved. It has just bought more of something that doesn’t convert.

Cost per click is useful for exactly one thing: understanding what the auction in your category looks like before you enter it. After that, stop looking at it.

What the benchmarks say, and how to use them

WordStream’s 2026 Google Ads benchmarks put legal services among the most expensive categories to advertise in, at roughly $9.87 a click and $131.63 a lead. Those are industry averages across many accounts, not our results and not a promise about yours.

Use a benchmark as a sanity check, not a target. If your click costs are a third of that figure in a legal category, the useful question is not "why are we so efficient" — it’s whether you are buying clicks from people who were never going to hire a lawyer. Benchmarks are most valuable when your numbers are far away from them and you want to know why.

Legal services sits among the highest-cost categories, at roughly $9.87 per click and $131.63 per lead.

WordStream, 2026 Google Ads benchmarks

Cost per lead is the number that misleads

Cost per lead feels like the honest number because it counts a human being who raised a hand. The trouble is that "lead" is not a standard unit. One firm counts every form fill. Another counts only calls that lasted more than a minute. A third counts booked consultations.

So when someone tells you their cost per lead is low, you have learned nothing until you know what they are counting. And when a campaign’s cost per lead improves, that improvement can be entirely made of leads that were never going to buy — price shoppers, people in the wrong city, people who wanted something you don’t sell.

The rule of thumb: any change in cost per lead that isn’t accompanied by a change in customers is not a real change.

Cost per customer is the only one that decides anything

Total spend divided by customers won. That is it. It is the only one of the three that sits in the same units as your revenue and can therefore be compared to it.

The reason it gets skipped is that it’s harder to get. It requires knowing which customers came from which source, which requires somebody writing that down, which is exactly the discipline most small businesses don’t have. So people report the number they can get instead of the number they need.

1. Pick a closed month — one far enough back that its leads have had time to become customers or not.
2. Total everything you paid that month: ad spend plus any management fee.
3. Count the customers you can trace to that campaign, honestly. If you can’t trace them, that is your first finding.
4. Divide. That is your cost per customer.
5. Compare it to what a customer is actually worth to you over their whole relationship, not just the first invoice.

Lifetime value is what makes a scary number fine

A cost per customer that looks alarming in isolation is often perfectly sensible once you know what a customer is worth. If your average customer spends a few hundred dollars once, a high acquisition cost kills you. If they stay for years, refer people, and come back, the same number is a bargain.

This is why click costs in legal, insurance and high-end home services get so high. Bidders are not being reckless. They are pricing against the whole relationship, and if you price against the first transaction you will be outbid by people doing arithmetic you aren’t doing.

It is also the honest argument for spending on retention. Anything that makes an existing customer stay longer or refer more improves the maths on every ad you will ever run, retroactively.

The leads you already paid for

Here is the thing that changes the arithmetic most and costs the least. Most businesses have a pile of leads they already paid to acquire and never followed up properly. Those are sunk cost. Converting one of them costs nothing in ad spend.

A campaign whose cost per customer looks bad often has a perfectly good cost per lead and a follow-up process that loses most of them. Fixing that lowers cost per customer without touching the ad account at all, because the denominator goes up while the spend stays the same.

Before increasing a budget, always check whether the existing leads are being worked. It is the cheapest improvement available and it is available to almost everybody.

The number between lead and customer

There is a fourth figure worth tracking, and it explains most of the gap between a good cost per lead and a bad cost per customer: your close rate. What share of inquiries become paying customers.

It matters because it multiplies. If your cost per lead is two hundred dollars and you close one in four, a customer costs eight hundred. Close one in two and the same leads produce customers at four hundred. You have halved your acquisition cost without changing a single thing in the ad account.

That is why sales process work is usually the cheapest lever available to a small business. Everyone reaches for the budget dial first because it’s the one on the dashboard. The close rate is the one that moves the number that matters, and improving it costs time rather than money.

It also tells you where to look when things go wrong. If cost per lead is stable and cost per customer is climbing, the ads are fine and something changed in how you handle inquiries.

Attribution will never be perfect, and that’s survivable

Somebody sees your ad, doesn’t click, looks you up two weeks later, reads a review, asks a friend, and calls. Which channel gets the credit? Any answer is partly arbitrary, and anyone who tells you their attribution is exact is selling something.

You don’t need exact. You need consistent and directionally honest. Ask every caller how they found you and write the answer down. Compare inquiry volume in months you advertised against months you didn’t. Those two crude methods, applied steadily, will beat elaborate dashboards built on tracking nobody verified.

What consistency buys you is the ability to compare months to each other. If you count the same way every month, even an imperfect method produces a trend you can trust, and a trend is what decisions actually get made from. Changing how you count halfway through the year destroys that, and it is a more expensive mistake than counting crudely in the first place.

The thing to avoid is letting the difficulty become a reason to stop measuring. "We can’t track it perfectly" turns quietly into "we don’t track it", and then every budget conversation for the next three years is one person’s hunch against another’s.

What to actually report every month

If you take one thing from this: change what you ask for. Most monthly reports are a wall of impressions and clicks because those are easy to produce and hard to argue with.

• Total spend, including management fee.
• Leads, with a stated definition that doesn’t change between months.
• Customers won that trace back to the campaign.
• Cost per customer, and how it compares to a customer’s worth to you.
• One thing that changed and what it did.