Resource

·

Acquisition

Cost per qualified lead

Cost per qualified lead: a floor of tiles seen from a low angle.

The short answer

Cost per qualified lead is acquisition spend divided by the number of leads that met a written definition, set by whoever takes the first call. It is the only paid metric that survives contact with a sales team, because cost per lead rewards whichever channel sends you the most people, and that is usually the channel sending the wrong ones.

By

Rohan

·

6 min read

Published

Oct 2, 2026

·

Updated

Oct 2, 2026

The formula

Cost per qualified lead = all acquisition spend in the period, divided by the number of leads created in that period that met the definition.

All spend means media, plus the tools and people you would stop paying for if you stopped acquiring. Leads created in the period, not leads qualified in the period, or a month with a slow sales week will look like a month with bad ads.

How do you define qualified so the number holds?

Four checkable fields, written with the person who takes the first call, in one place.

  • Fit: rules a machine could apply. An employee band, a revenue band, a country list, a role category, an industry exclusion list. No judgment calls.

  • Need: their own answer to one open question on the form, in their words.

  • Timing: a stated window. This quarter and sometime next year are different leads.

  • Access: whether they can get the signer into a room, not whether they are the signer.

Then test it. Take the last 20 leads, give the list separately to the founder, the person who takes first calls, and whoever runs marketing, and have each mark qualified or not with no discussion. Fewer than 17 of 20 in agreement means the definition is still a mood. Rewrite the rule that caused each disagreement and run it on 20 more.

Store it as one field on the record in the CRM, set by the person who takes the call, with a short list of reasons when the answer is no. The reason codes are the cheapest reporting you will ever build: three months of "wrong country" and "no budget this year" tells you which campaigns to switch off without touching an attribution model.

What should yours be?

Published averages will not tell you. Across thousands of Google Ads and Microsoft Ads accounts, the average cost per lead was $66.69, with a range from $26.84 in arts and entertainment to $131.63 in legal services (LocaliQ, Search Advertising Benchmarks, updated June 1, 2026). A five-fold spread inside one dataset, and none of those leads was tested against your definition.

Work from your own margin instead. Five lines:

  1. First-year revenue per customer.

  2. Gross margin on it.

  3. The payback period your cash can carry.

  4. CAC ceiling = annual gross profit per customer × (payback months ÷ 12).

  5. Cost per qualified lead ceiling = CAC ceiling × the share of qualified leads that become customers.

Worked example. A $12,000 contract at 80% margin leaves $9,600 a year. A 12-month payback gives a CAC ceiling of $9,600. If one in five qualified leads closes, the ceiling for a qualified lead is $1,920. If one in three form fills clears the definition, the raw cost per lead ceiling is $640, roughly ten times the published average above.

That gap is the reason to do the arithmetic. The benchmark says $66.69, your ceiling says $640, and a $200 cost per lead is either comfortable or ruinous depending on a close rate no benchmark knows.

Two cautions. The ceiling is a ceiling: paying it exactly means a year of growth that earns back only what it cost. And if you have fewer than about 30 closed-won deals, your close rate has an error bar wide enough to move the ceiling by half, so set the ceiling from the lower end of what you have seen and revisit it every quarter.

Where does the number live?

One report, in one place, showing per channel and per campaign: spend, leads, qualified leads, cost per qualified lead, and the share that qualified. The last column is the one that catches a channel going bad early, because quality falls before cost rises.

For the number to trace back to a campaign at all, the lead's origin has to be captured at the moment the form is submitted, as hidden fields written to the CRM record. It cannot be reconstructed later. Attribution without the arguments has the field list and the convention.

How do you get the qualified event back to the ad platforms?

The platforms optimize toward whatever you send them, so a funnel that only reports form fills will be fed more form-fillers. Both large platforms accept a later event from your own systems: Google Ads through offline conversion imports, Meta through its Conversions API, each matched to the click or the lead using the identifier you captured at submit. This is the reason the click ID belongs in the hidden fields on the form.

Two conditions before you do it. The qualified event has to happen often enough to be worth optimizing toward, which brings you back to the volume arithmetic: below roughly 50 of them a week per ad set on Meta, the platform cannot learn from them. And the event has to be recorded consistently, because an inconsistent signal sent back at scale trains the platform on your bad month.

If the volume is not there, send the event anyway for reporting and keep optimizing toward an earlier step, then check quality by hand every week.

The five mistakes

  1. Reporting cost per lead instead. It makes the channel with the loosest form look like the best channel in the account.

  2. Letting marketing define qualified. The definition has to be owned by whoever will be judged on calling those people, or it will be relaxed the first slow month.

  3. Counting qualified at the wrong moment. Qualification happens on the first call. If it is set by a scoring rule at form submit, you are measuring your own guess.

  4. Optimizing the ad platform toward the raw form fill. The platform will then find you more people who fill in forms. If the qualified event can be sent back to the platform, send it; if its volume is too low to optimize toward, optimize toward an earlier event and police quality by hand each week. The arithmetic for that choice is in the creative testing loop.

  5. Comparing this month to last month without a sales-cycle check. A month with a holiday in it produces fewer first calls, so fewer leads get qualified, and the cost per qualified lead rises for a reason that has nothing to do with acquisition.

The one-week checklist

  • Monday: write the four fields with the person who takes first calls. One page.

  • Tuesday: run the blind test on the last 20 leads. Rewrite what disagreed.

  • Wednesday: add the qualified field and the reason codes to the CRM. Add the hidden fields to every form.

  • Thursday: do the five-line ceiling arithmetic. Write the number on the same page as the definition.

  • Friday: build the one report, with the share-that-qualified column, and put it on the agenda of a meeting both teams already attend.

How I use it

With a Series A B2B SaaS company whose paid spend had grown for three quarters against flat demo volume, this was the first week of work, before anything in the ad accounts changed.

The blind test was the part that moved people. The founder, the sales lead and the marketer disagreed on 7 of 20 leads, and almost every disagreement was about company size, which everyone had assumed was settled because they all used the word "mid-market". We replaced the word with an employee band and a revenue band, ran it again on a fresh 20, and disagreed on two.

Then the ceiling arithmetic, which produced a number well above what they were paying, and that was the surprise in the other direction: their cost per lead was fine and their qualification rate was terrible. Half the demos were not worth taking, so the real cost of a conversation worth having was roughly double what anyone had been quoting in board meetings.

What I would do differently: the first version of the definition included a "level of interest" field, which nobody could score the same way twice. It got argued about for a month before we deleted it. I now refuse any field a stranger could not fill in identically on two different days.

Takeaways

Write the four fields with the person who takes the first call, then blind-test them on 20 recent leads.

Compute your ceiling from margin and close rate, and write it on the same page as the definition.

Put the share-that-qualified column next to cost in the report, because quality falls before cost rises.

Related

Questions

What is a good cost per qualified lead?

The one your margin can carry, which you compute from first-year gross profit, your payback period and your close rate. Published averages span a five-fold range by industry and none of those leads was tested against your definition.

Who should define qualified, marketing or sales?

How often should the definition change?

Sources

If you’d like to talk through how this applies to your company, I’d be glad to.