Cost per lead (CPL) is ad spend divided by leads, and cost per sale (CPS) is ad spend divided by closed sales; CPL tells you how cheaply you fill the funnel, while CPS tells you what revenue actually costs, so CPS is the number to optimize whenever a sales team sits between the lead and the money. The two are linked by one formula: cost per sale = cost per lead ÷ lead-to-sale conversion rate. A campaign with a $10 CPL and a 2% close rate costs $500 per sale. A campaign with a $40 CPL and a 12% close rate costs $333 per sale. The "expensive" leads win.
This guide explains when each metric is useful, shows the math with a worked example, and gives you a step-by-step way to start reporting cost per sale by campaign.
What is the difference between cost per lead and cost per sale?
| Cost per lead (CPL) | Cost per sale (CPS) | |
|---|---|---|
| Formula | Spend ÷ leads | Spend ÷ closed sales |
| Where the data lives | Ad platform or form tool | CRM (closed-won deals) |
| How fast you see it | Same day | Days to months, depending on sales cycle |
| What it rewards | Volume of contacts | Contacts who buy |
| Typical failure | Cheap leads that never answer the phone | Too few sales to read per campaign |
| Best used for | Early signal, creative tests, capacity planning | Budget decisions, channel and campaign ranking |
Cost per sale is sometimes called cost per acquisition. When "acquisition" means a new paying customer, it is the same as CAC by channel calculated on media spend only.
What are the formulas, and how do they connect?
- CPL = ad spend ÷ number of leads
- Cost per qualified lead = ad spend ÷ leads that pass your qualification rule
- Cost per sale = ad spend ÷ number of closed sales
- Cost per sale = CPL ÷ lead-to-sale rate
The last line is the useful one. It shows that you can lower cost per sale in two ways: buy leads more cheaply, or buy leads that close more often. Ad platforms only see the first lever unless you send them sales data. That is why optimizing on CPL alone often makes cost per sale worse.
What does a worked example look like?
Example (hypothetical numbers): two Meta campaigns, same month, same budget, same sales team.
| Metric | Campaign A (instant form) | Campaign B (website form + qualifying question) |
|---|---|---|
| Spend | $5,000 | $5,000 |
| Leads | 500 | 125 |
| CPL | $10 | $40 |
| Lead-to-sale rate | 2% | 12% |
| Sales | 10 | 15 |
| Cost per sale | $500 | $333 |
| Revenue (avg. deal $1,200) | $12,000 | $18,000 |
| Sales hours (15 min per lead) | 125 hours | 31 hours |
A media buyer looking only at CPL would move budget to Campaign A. That decision would cut sales by a third and quadruple the hours your sales team spends calling people who will not buy. The hidden cost of cheap leads is sales capacity. If a rep costs $25 per hour, Campaign A's real cost per sale rises to about $813, against roughly $385 for Campaign B.
Why does cost per lead mislead so often?
- The algorithm finds the cheapest form fills. If you tell Meta or TikTok that a lead is the goal, they will find people who submit forms, not people who buy.
- Low-friction formats inflate volume. Pre-filled instant forms and "Hi" messages on WhatsApp are easy to send by accident or out of curiosity.
- Duplicates and junk count as leads. The same person submitting twice, fake numbers and bots all lower CPL.
- Lead definitions drift. A form fill, a WhatsApp conversation and a booked call are all "leads" in some reports.
CPL is a vanity metric when it is the only thing you report. Our post on vanity metrics vs revenue metrics covers the wider pattern.
When is cost per lead still the right metric?
- Early creative tests, when you need a read within days and sales data will take weeks.
- Capacity planning, to know how many leads your team can handle.
- High-volume, short-cycle businesses where lead quality is stable and close rates barely vary by campaign.
- As a leading indicator, always shown next to cost per qualified lead and cost per sale.
Which funnel metrics should you track between lead and sale?
| Metric | Formula | Speed | Use it to |
|---|---|---|---|
| Cost per lead | Spend ÷ leads | Same day | Spot broken ads quickly |
| Cost per qualified lead | Spend ÷ qualified leads | 1 to 3 days | Judge lead quality before sales close |
| Cost per opportunity | Spend ÷ opportunities or quotes | 1 to 3 weeks | Read mid-funnel in long cycles |
| Cost per sale | Spend ÷ closed sales | Sales cycle length | Rank campaigns and channels |
| ROAS or ROI | Revenue ÷ spend, or (revenue − cost) ÷ cost | Sales cycle length | Decide how much to scale |
If your sales cycle is long, cost per qualified lead is the best early proxy. It only works if "qualified" is a fixed rule in your CRM, such as "budget confirmed and decision maker reached," not a rep's mood.
How do you start measuring cost per sale by campaign?
- Capture the source on every lead. UTMs, click IDs and the WhatsApp ad referral must be saved at the first touch.
- Send that source into the CRM with each contact, through hidden form fields or an integration. See our lead tracking page for how this works across forms and WhatsApp.
- Define the sale stage. Pick one CRM stage, such as "closed-won" or "first payment," and never change it mid-quarter.
- Join spend and sales by campaign in one table. Spend comes from the ad platforms, sales from the CRM.
- Report CPL weekly and cost per sale monthly, and use cohorts if deals take more than a few weeks to close.
- Send sales and qualified stages back to the ad platforms so they can optimize for buyers, not form fills. On Meta this is the conversion leads goal, covered in how to optimize Meta ads for qualified leads. TikTok has a similar deep funnel optimization for lead generation that uses CRM events.
Step 6 matters because it changes what you buy, not only what you report. Our guide to connecting offline CRM conversions to ad platforms explains the data flow.
Where does Admira fit?
Admira ties Meta, Google Ads, TikTok and Click-to-WhatsApp leads to closed deals in your CRM, so you can see cost per lead, cost per qualified lead and cost per sale per campaign in one report. Its Conversion API sends CRM stages back to ad platforms server-side. It is not the only way to do this, and it is a young brand with few public reviews. If you sell through an online checkout with no sales team, an ecommerce attribution tool is likely a better fit.
Who should focus on cost per sale, and when doesn't it apply?
Focus on cost per sale if a person (sales rep, advisor, WhatsApp agent) follows up on leads, your close rate differs by campaign, or your deals are worth enough that a few extra sales change the month. Real estate, education, lending, B2B services and clinics are typical. For a sector example, read how to measure cost per sale in real estate.
It applies less when the lead is the product (for example, you sell leads to partners), when you close fewer than a handful of sales per campaign per month (use cost per qualified lead instead), or when your sales cycle is so long that cost per sale arrives after budget decisions are made.
FAQ
Is cost per sale the same as CAC?
Close, but not always. Cost per sale usually counts media spend only and can include repeat sales. CAC counts only new customers and may include salaries and tools.
What is a good cost per lead?
A good CPL is one that produces an acceptable cost per sale. Work backward: target cost per sale times your expected close rate gives your maximum CPL.
How do I calculate the maximum CPL I can afford?
Multiply your target cost per sale by your lead-to-sale rate. With a $400 target and a 5% close rate, the maximum CPL is $20.
Why did my CPL drop but sales stayed flat?
Usually because the platform found cheaper, lower-intent leads. Check close rate by campaign before celebrating a lower CPL.
Can I see cost per sale inside Meta Ads Manager?
Only if you send sales from your CRM back to Meta through the Conversions API. Even then, compare it with your CRM, because platforms can count a sale differently.
How many sales do I need to compare campaigns?
There is no fixed cut-off, but with fewer than about 10 sales per campaign, differences are often noise. Use longer windows or compare cost per qualified lead.
.svg)
%20(2).png)
