Skip to content
  1. Home
  2. Glossary
  3. Cost per qualified lead
Spend & efficiency

Cost per qualified lead

What you pay for a lead the sales team actually wants — and why it is the first number that tells you whether cheap traffic is cheap.

Formula
spend / qualified leads
Unit
Currency
Direction
Lower is better
Across periods
Recomputed from totals, never averaged
In KO
cpQl

Cost per qualified lead is what you paid for each lead that passed qualification — that survived first contact and turned out to be a real prospect with a real need.

It exists because cost per lead (CPL) can be driven down indefinitely by buying worse leads, and nothing in CPL will object.

The formula

CP QL = spend / number of qualified leads

€10,000 spent, 400 leads, of which 120 qualify:

CPL   = 10,000 / 400 = €25
CP QL = 10,000 / 120 = €83

The gap between the two is the price of unqualified volume. A channel with a €25 CPL and a 30% qualification rate is more expensive per real prospect than a channel with a €40 CPL that qualifies at 70% — €83 against €57.

This is the whole argument for the metric: CPL ranks channels by what they cost, CP QL ranks them by what they are worth. The two rankings routinely disagree, and the disagreement is the finding.

What it does not show

It does not carry your definition of “qualified” to anyone else. Qualification is a rule you wrote — budget confirmed, decision-maker reached, ICP match, a stage in your CRM. Two companies with identical funnels and different rules produce CP QL figures that cannot be compared. Neither can two teams inside one company, if sales changed the rule in March. There is no industry benchmark for CP QL, and any published one is measuring someone else’s definition.

It does not know whether qualified leads close. A channel can produce cheap, genuinely qualified prospects that never buy — wrong company size, wrong timing, price shock at the proposal. CP QL is the last purely marketing-side metric in the chain; everything after it is shared with sales.

It says nothing about capacity. Halving CP QL doubles the qualified leads at the same budget, which is only good news if someone can work them. Beyond the team’s capacity, the extra leads age in the pipeline and convert worse — the metric improves while the outcome does not.

It hides its own denominator’s delay. Qualification happens days or weeks after the lead arrives. Recent periods always look expensive because their leads have not finished qualifying yet, and the number improves on its own as the backlog clears. Comparing this week to last quarter compares a partial denominator against a settled one.

How it breaks

Attributing spend and leads to different periods. Spend is recorded on the day of the click; the lead may qualify three weeks later. If spend is bucketed by click date and qualified leads by qualification date, the two sides of the ratio describe different cohorts. It is a cohort question, and reporting it as a period question quietly mixes them.

Averaging it across periods or channels. Like every ratio, CP QL must be recomputed from totals — Σspend / Σqualified — not averaged from the level below. Averaging per-channel CP QL weights a channel with 3 qualified leads the same as one with 300.

Unattributed leads shrinking the denominator. Qualified leads whose source was never recorded do not belong to any channel, so every channel’s CP QL looks worse than it is — you divide the full spend by a fraction of the results. When a meaningful share of leads has no source, per-channel CP QL is not wrong so much as unusable, and the fix is the tracking, not the metric.

A qualification rule that moved. If sales tightened the criteria, CP QL rises with no change in marketing. This is the most common false alarm in the metric, and it is invisible in the number itself — the only trace is in the CRM’s stage history.

How KO calculates it

cpQl is a derived metric: spend / qualified, recomputed from summed inputs at every level of a report rather than averaged from the rows beneath it. Where the qualified count is zero, the result is reported as empty (—), not as zero or a very large number.

What counts as qualified comes from your CRM’s stage mapping, not from a KO default — which is also why KO flags projects where no stage has been classified as qualification: until that mapping exists, everything downstream of it, including this metric, is uncomputable rather than merely inaccurate.