How do I know if paying for leads is profitable?
Short answer
Work backward from the lead price to the cost of a customer. A $20 lead you reach half the time is $40 a conversation. Close one in four and a customer costs about $160. Against a $1,500 average sale that's roughly 10X, and e-commerce operators scale hard at 2X. Then run a short test before you trust the numbers.
Most people look at the cost per lead and panic, or they look at it and celebrate. Neither tells you anything. What matters is what a customer costs you compared with what a customer is worth.
How do I get from cost per lead to cost per customer?
Walk it backward. Here's the example I used with a business owner, with easy numbers:
| Step | Number |
|---|---|
| Cost per lead | $20 |
| Leads you actually reach | 1 in 2 |
| Cost per conversation | $40 |
| Close rate on conversations | 25% (1 in 4) |
| Cost per customer | $160 |
"If a conversation is costing you $40 and you have to talk to four of them, that's $160."
Note the close rate is on people you speak to, not on leads. The business owner I was helping said about half her website leads pick up and she closes three or four out of ten she talks to, so my 25% was conservative for her.
What's a customer worth?
Use your real average, not your best or worst deal. Take your last 30 closings, add them up and divide by 30. You don't care that one time you only made $400, because another time you'll make $2,500.
Then add lifetime value. If maybe half the people you help come back next year, a $1,500 client is really worth something like $2,200.
What return should I want?
If you make $1,500 on an average sale and a customer costs you $160, that's about 10X. E-commerce companies have a term for what they do at 2X: back up the truck. They put in as much money as they can. At 10X, you should be spending more.
The bigger the sale, the more you can pay for a lead. On another call, the math was a $300 lead with one in four closing. That's $1,200 to win a client worth $20,000 in revenue. You'd do that all day.
How long do I test before I trust it?
The risk is spending $200, $300, $400, $500 before the first lead and having nothing for it. Maybe it won't work at all. It's still worth trying.
- Run a short test, about two or three weeks.
- Even one or two weeks of results can show you it's clearly going to work.
- Keep tracking how many leads you reach and how many you close, so your numbers are real, not guesses.
If you're paying for newsletter subscribers instead of sales leads, the same thinking is in how much you can pay per subscriber. If you're the one selling leads, see how much to charge per lead.
Answered by TJ Larkin. From consulting calls with business owners, 2025-2026. Last updated .