Someone tells you $40 a lead is good. They have no idea what you sell, what it earns you, or how many enquiries turn into customers. A good cost per lead cannot be handed to you by a stranger, because half of the calculation lives inside your own business.
The short answer: your close rate and your profit per sale set the ceiling, and everything below that ceiling is fine. That is why two companies in one industry can have targets that differ by a factor of ten and both be correct.
This is for you if you are trying to work out whether to keep spending, spend more, or stop.
Why there is no universal good cost per lead
A published average tells you what other advertisers paid. It cannot tell you what you can afford, because affordability depends on numbers that never appear in a benchmark table.
For scale, WordStream’s 2026 study of 13,474 US search campaigns put the all-industry average at $66.69, with an average conversion rate of 8.18%. That is genuinely useful for knowing whether you are in a cheap market or an expensive one.
It becomes harmful the moment you treat it as a goal. Chasing a number set by other people’s economics is how businesses either overspend on leads they cannot monetise, or throttle campaigns that were making money.
The three numbers that decide your target
You need profit per sale, close rate, and the margin you want to keep.
Profit per sale means what you actually keep, after cost of delivery. Not revenue. Using revenue here is the single most common error, and it produces a ceiling two or three times too high.
Close rate means the share of leads that become paying customers. Not the share that reply, not the share that book a call. Paying.
Margin is your buffer. Break-even is where lead value equals lead cost, and no business wants to operate there.
Working out your maximum cost per lead
Multiply profit per sale by close rate. That gives you the value of one lead, which is your break-even ceiling.
A business earning $3,000 profit per sale that closes one lead in twenty has a lead value of $150. Break-even is $150. Anything below makes money.
Now change one input. Keep the same $3,000 but close one in ten, and lead value doubles to $300. Same product, same price, twice the acceptable cost, because the sales process got better.
That is the point worth sitting with. Improving your close rate raises what you can afford to pay, which lets you outbid competitors who have not done that work. Nothing you change inside an ad account has that effect.
Most people then set their working target below break-even, often around half. That buffer covers the leads that take three months, the ones that ghost, and the months where the auction is expensive.
A good cost per lead can be a large number
I have run an account at $91.18 per lead that was one of the better performers on my books.
It was 906 leads for a US automobile accessories business on Google, in a market where a single closed order is worth thousands. Set against a benchmark table, $91.18 looks like something went wrong. Set against what the client earned, it looks like a bargain.
In the same period I ran a US auto transport account on Meta at $2.68 per lead across 5,599 leads. Two completely different figures, both correct for the business behind them.
A $91 lead is not expensive if it is worth $4,000. A $9 lead is expensive if it never answers the phone.
Your good cost per lead differs by channel
Search and social are not buying the same person, so holding them to one number hides what is happening.
Search catches someone already looking for what you sell. Social interrupts someone who was not thinking about you at all. The first tends to close at a higher rate, which means it justifies a higher price per lead.
Ruler Analytics, working from 110 million sessions across 13 industries, measured paid search converting at 5.4% against 2.11% for paid social. Roughly two and a half times the rate.
So if your social leads cost half as much as your search leads, that is not automatically a win. Priced per closed customer, they may be the more expensive channel.
Signs your target is set wrong
Four symptoms come up again and again, and each points at a different mistake.
Your campaigns look profitable but the bank balance disagrees. That usually means the ceiling was built on revenue rather than profit, so every lead is bought above its real worth.
You keep pausing campaigns that were working. That is a target set at break-even with no buffer, so ordinary monthly variation reads as failure.
Sales complain about lead quality while marketing celebrates the cost. That is a target chased without watching who is actually arriving, which is the easiest way to lower a figure and damage the business at the same time.
Nothing has ever hit the target, in any month, on any channel. At that point the number is not a target, it is a wish. Either the economics do not support paid advertising yet, or the ceiling was copied from somebody else’s business.
That last one is worth taking seriously rather than pushing through. Advertising cannot fix a product that does not make enough money per sale.
What to change this week
Four steps, in order.
Pull the last 90 days of leads and count how many became customers, so you have a real close rate rather than a hopeful one. Work out profit per sale, using profit and not revenue. Multiply the two to get lead value, then set your working ceiling at roughly half of it. Split that ceiling by channel rather than running one number across everything.
If your figure climbs the moment you raise budget, that is normal auction behaviour with its own causes and fixes. The broader piece on judging this metric covers where it fits with everything else, the maths behind the calculation covers the inputs, and the numbers from accounts I have run show what the range looks like in practice. If you want a second opinion on your ceiling before you set budgets, tell me what you sell.