Cost per lead or cost per sale: which number to watch in paid ads

Short answer: cost per lead is useful for day-to-day campaign monitoring, but cost per sale is what tells you whether paid ads are profitable. A campaign with cheap leads can lose money, and one with expensive leads can be the one that sells the most. That is why management has to tie every lead to the sale it generated and set the budget by the number that pays the bills.

In this article

  • What cost per lead and cost per sale are
  • Why cheap leads are misleading, with a worked example
  • How to find the maximum cost per sale your business can afford
  • Step by step to measure the sales of each campaign
  • How to use both numbers together in your routine
  • Frequently asked questions

What cost per lead is

Cost per lead (CPL) is how much you spent on ads divided by the number of contacts that came in: forms submitted, WhatsApp conversations, phone calls or Instagram DMs. If you spent US$ 3,000 in the month and got 300 leads, your cost per lead was US$ 10.

It is a useful number because it shows up fast, right inside the ads manager, and tells you whether the campaign is bringing people in. It also reacts quickly when something breaks: the creative wore out, the form stopped working, the audience got saturated. The problem starts when it becomes the main goal of the account.

What cost per sale is

Cost per sale, also called cost per acquisition (CPA), is your ad spend divided by the number of sales those ads generated. In the same example, if 15 sales came out of those 300 leads, your cost per sale was US$ 200.

With that number you can answer the question that matters to the business owner: did each sale cost less than the margin it left? If your average ticket is US$ 1,500 and your margin is 40%, each sale leaves US$ 600. Paying US$ 200 for it is a good deal. Paying US$ 700 is not, no matter how cheap the lead looks.

There is a third number that closes the loop: return on ad spend (ROAS), which is the revenue generated divided by the amount spent. It works great for ecommerce, where the sale happens on the website. For businesses that sell over WhatsApp, phone or in meetings, cost per sale is usually easier to understand and to compare across campaigns.

Why cheap leads are misleading

Ad platforms learn from what you send them. If a campaign is optimized for leads, the algorithm will look for the people most likely to fill out a form, and those are not always the people who buy. Instant forms inside Facebook, for example, tend to bring cheaper and colder leads than a form on your own page, because they are too easy to submit.

Campaign ACampaign B
SpendUS$ 2,000US$ 2,000
Leads400120
Cost per leadUS$ 5US$ 16.67
Sales412
Close rate1%10%
Cost per saleUS$ 500US$ 166.67

Looking only at cost per lead, campaign A seems three times better. Looking at sales, campaign B delivers three times the result with the same money. Whoever turns off B because its leads are expensive turns off the very campaign that sells.

This scenario is very common. The most frequent reasons cheap leads do not buy are:

  • Curious audiences: people who click on everything and fill out forms on impulse.
  • Misaligned promise: the ad talks about low prices and the product is premium.
  • Forms that are too easy: no question that filters for real interest.
  • Wrong region or profile: leads from cities you do not serve, or consumers when you sell to businesses.
  • Slow response: the lead was good, but went cold because nobody answered in time.

Notice that only the last item is a sales problem. The other four are campaign problems that cost per lead hides.

The maximum cost per sale your business can afford

Before judging whether a campaign is good, you need to know how much you can pay for a sale. The math starts with the contribution margin of each sale, that is, what is left after product or service costs, taxes and commissions.

  1. Take your average ticket. Example: US$ 1,500.
  2. Subtract costs, taxes and commission. If they add up to 60%, you are left with a US$ 600 margin.
  3. Decide how much of that margin you are willing to spend on acquisition. Many companies work with somewhere between a third and a half. With half, your ceiling is US$ 300 per sale.
  4. Consider repeat purchases. If customers usually buy again, a customer is worth more than the first order, and the ceiling can go up.

With that ceiling in hand, the reading becomes objective: below US$ 300 per sale, the campaign deserves more budget; above it, it needs adjusting or cutting. Without a ceiling, the discussion turns into opinion.

How to measure cost per sale in practice

The math is simple. The hard part is having the right data. The path we use with our clients has five steps.

1. Tag the source of every lead

Every contact needs to arrive with its source recorded: platform, campaign, ad set and ad. You do this with UTM parameters on your ad links, with click identifiers (Google's gclid and Meta's fbclid) and with tracking set up on the website to keep that information until the moment the form is submitted.

2. Record the lead in the CRM

Loose spreadsheets and conversations lost on WhatsApp break the link between the ad and the sale. The lead has to go into the CRM automatically, with the source already filled in, without depending on someone copying and pasting. We explain how in the article on CRM for small businesses.

3. Mark the sale on the same record

When the lead buys, the salesperson moves the deal to "won" in the CRM and enters the value. This is the step that fails most often, because it depends on team discipline. A simple pipeline, with few stages and required fields only where they matter, helps a lot.

4. Cross spend and sales by source

With source and sale in the same place, cost per sale comes out per campaign and even per ad, not just as a monthly total. That level of detail is what lets you move budget with confidence.

5. Send the sale back to the platforms

The step that changes results the most is sending the sale back to Meta and Google. When a platform knows which leads became customers, it starts looking for people similar to buyers, not to form fillers. On Meta this is done with the Conversions API, sending the purchase event with the lead's data protected by hashing. On Google Ads, with offline conversion imports or enhanced conversions for leads. We explain how it works in the article on the Meta Conversions API.

How long it takes for the sale to show up

An important caveat: the sale does not always happen on the same day as the lead. In high-ticket services the cycle can take weeks. If you calculate last week's cost per sale, you will think the campaign is terrible when the sales are actually still in the pipeline.

So look at cost per sale in windows that match your sales cycle. If most sales close within 30 days, compare one month's spend with the sales generated by that month's leads, even if some close the following month. A good CRM does this math by the date the lead came in, not by the date of the sale.

How to use both numbers together

Cost per lead does not die. It becomes the thermometer, and cost per sale becomes the compass. The routine we recommend looks like this:

FrequencyNumberTo decide
Daily or every 2 daysCost per lead, spend, deliveryWhether something broke: tired creative, form down, stuck budget
WeeklyLead qualification rateWhether the audience and the ad promise attract the right profile
Every two weeks or monthlyCost per sale and return by campaignWhere to add budget, what to cut and what to test

A practical rule: never turn off a campaign just because cost per lead went up. First check whether quality went up with it. Very often the lead got more expensive because it got better.

Common mistakes that distort cost per sale

  • Counting sales from existing customers as campaign sales. Someone who was already a customer and clicked an ad inflates the result. Separate new and returning customers.
  • Ignoring sales that arrive through another channel. The lead sees the ad, disappears and comes back through Google a week later. Without the source saved at first contact, the credit goes to the wrong place.
  • Counting duplicate leads. The same person fills out the form twice and counts as two leads.
  • Forgetting costs outside the platform. Tools, creative production and agency fees are part of the total acquisition cost, even if they are not part of the campaign's cost per sale.

Frequently asked questions

What is a good cost per lead?

There is no universal number. A US$ 80 cost per lead can be great for a US$ 20,000 service and terrible for a US$ 150 product. What defines whether a lead is expensive is the cost per sale it generates, compared with your margin.

What if I do not have a CRM yet?

You can start with a well-organized spreadsheet, with the source of each lead and a sales column. It works for a few weeks. But once volume passes a few dozen leads a month, the spreadsheet starts losing information, and a CRM pays for itself.

Should I optimize my campaign for leads or for sales?

If you can send sales back to the platform with enough volume, optimizing for the stage closest to the sale usually works better. If you cannot yet, optimize for leads, but track sales separately and use that data to choose campaigns and audiences.

How soon can I trust cost per sale?

It depends on volume. With few sales per month, each extra or missing sale changes the number a lot. Prefer 30 to 90 day windows and compare trends, instead of reacting to a single week.

Summary

  • Cost per lead measures contacts. Cost per sale measures results.
  • Calculate your cost per sale ceiling from your margin before judging any campaign.
  • To measure sales, the lead's source and the sale must live on the same CRM record.
  • Respect the sales cycle: compare leads from a period with the sales they generated.
  • Sending sales back to Meta and Google teaches the platforms to find buyers.