Cost per lead is the wrong metric: how to measure advertising in sales
Why cost per lead misleads B2B companies, and how to measure every campaign in qualified leads and closed sales, with real numbers.
A campaign with a €25 cost per lead looks better than one with a €1,400 cost per lead. In a lot of B2B companies, it is not. The cheap campaign brings forms. The expensive campaign brings buyers. If cost per lead is all you look at, the budget ends up in the wrong campaign.
This article explains why that happens, what to measure instead, and what it takes to get there.
What cost per lead measures, and what it does not
Cost per lead divides spend by the number of forms received. It is easy to calculate and it appears in any Google or Meta dashboard. That is how it became the default metric.
The problem is what it cannot see:
- It does not tell a lead asking for a quote apart from someone looking for a job.
- It does not know whether a salesperson ever spoke to that person.
- It does not know whether there was a sale, or how much it was worth.
- It does not know whether the customer is still there three months later.
In B2B, the distance between the form and the sale is long. Weeks, sometimes months. And it is in that distance that the information you actually need gets lost.
A real case: €25 against €1,400
Aritco manufactures home lifts and sells through a network of dealers. In its campaigns, cost per lead on Meta was €25. On Google, around €1,400.
On that number alone, the decision looks obvious: move budget to Meta.
Only the truth from the CRM tells you which of the two channels actually qualifies. A qualified lead is one that meets the criteria the company defines: they own a home, they need the product, they can pay for it. Once every lead is followed through to that point, the comparison changes meaning entirely. You are no longer comparing the price of a form. You are comparing the price of a likely buyer.
Without that number, the decision is made blind. With it, the decision is made with the P&L in front of you.
The four metrics that replace cost per lead
| Metric | What it answers | Where it comes from |
|---|---|---|
| Cost per qualified lead | What a contact worth calling actually costs | Ads + the qualification criteria in the CRM |
| Cost per sale | What it costs to close a customer | Ads + closed sales in the CRM |
| Attributed revenue | How much each campaign actually billed | The value of each sale joined to its origin |
| Return by campaign | Which campaign returns more than it costs | Attributed revenue / spend |
None of the four comes out of the ad platform. All of them need the ad and the CRM to talk to each other.
An example with the full set of numbers
Talent Search People is a recruitment consultancy. A search campaign with €397 of spend produced 12 leads. Of those 12, 9 were qualified. Of those 9, 3 turned into a sale. Revenue: €27,200.
Using cost per lead, the report would say: €33 per lead. A correct and useless number.
With full measurement, the report says: €44 per qualified lead, €132 per sale, €27,200 of revenue from €397 of spend. That is a number you can set a budget with.
Why most companies do not measure it
It is not for lack of interest. It comes down to three concrete obstacles.
The CRM and the ads live apart
The form reaches the CRM without knowing which campaign it came from. Someone would have to note it down by hand. Nobody does that consistently.
“Qualified” is not defined
Every salesperson has their own idea of it. Without a written definition applied inside the CRM, there is nothing to measure. Defining “qualified” is the first job, before any tool.
The sales cycle is long
This month’s sale came from a lead two months ago. If you compare September’s spend against September’s sales, the maths will not add up. You have to measure by cohort: each sale is assigned to the month its lead came in.
What changes when the signal goes back to the platform
Measuring properly has a second effect, less visible and more important. Google and Meta optimise towards whatever you tell them a conversion is. If the conversion is “form submitted”, they will go looking for people who fill in forms. If the conversion is “qualified lead” or “closed sale”, they will go looking for people who buy.
This is done by sending offline conversions (Google) or the CAPI signal (Meta) from the CRM. The platform learns from real buyers and stops learning from browsers.
At Imparagon, Nexus does that work: it connects the ads to the CRM, follows every lead through to qualified and won, and sends that signal back to Google and Meta. As of September 2026 it has followed more than 34,000 leads through to the sale and €18.7M of revenue attributed to campaigns.
What to do this week
- Write down your definition of a qualified lead. One sentence. Something any salesperson can apply.
- Check whether the CRM stores the source campaign for each lead. If it does not, that is the first thing to fix.
- Pull the last six months of sales and join them to the lead that started them. A spreadsheet is fine.
- Compare cost per sale across channels. The result will probably surprise you.
Before you change anything
Measuring in sales is not a change of tool. It is a change of criteria. It forces you to define what a good lead is, to join systems that do not currently talk to each other, and to wait for the full sales cycle before drawing conclusions.
If you want to know which parts of that work your company already has done and which are missing, the diagnostic reviews your accounts, your CRM and your definition of qualified, and comes back with what is there and what is not.