CPL is the single easiest marketing number to make look good, because it only measures the top of the funnel. A landing page offering a discount code will produce a lower CPL than a demo request form every time, and neither number says anything about which one produces customers.
As with CAC, there's a channel version and a blended version. Channel CPL divides ad spend by form fills. Blended CPL adds the landing-page tooling, the SDR time qualifying inbound, and any agency fee tied to lead generation specifically. Reporting only the channel number is how a campaign gets more budget for producing worse leads.
CPL means nothing without lead-to-customer rate sitting next to it in the same report. A channel with double the CPL but triple the close rate is the cheaper channel, and comparing CPL alone would recommend cutting it.
Formula
CPL = Total Lead-Generation Spend ÷ Leads Generated
Always pair it with lead-to-opportunity or lead-to-customer rate. CPL ranks channels correctly only when quality is held constant, which it rarely is.
Why Cost Per Lead matters
Ad platforms optimize toward whatever conversion event you feed them. Set the objective to "lead" with no downstream quality signal and the algorithm will find the cheapest possible form fills, which reliably correlates with the least sales-ready traffic. CPL going down while sales complains leads are worthless is the platform doing exactly what it was told.
The cheaper channel isn't the one with the lower CPL
Channel A spends $8,000 and generates 200 leads: a CPL of $40. Channel B spends the same $8,000 and generates 80 leads: a CPL of $100, looking two and a half times worse. But Channel A closes at 2% (4 customers, effective CAC of $2,000) while Channel B closes at 15% (12 customers, effective CAC of $667). The channel with the worse CPL produced three times more customers per dollar.
Common mistakes
Ranking channels by CPL alone
CPL with no close-rate context rewards whichever channel has the lowest-friction form, not the one that produces customers. See the worked example. The ranking flips entirely once quality is added.
Feeding platforms the lead event with no quality feedback
Optimizing to raw lead volume trains the algorithm to find cheap form-fills, not buyers. Feeding back qualified-lead or opportunity status as the true conversion event fixes what the platform actually optimizes toward.
Excluding SDR time and tooling from the number reported upward
Channel-only CPL understates the real cost the same way channel-only CAC does. It just hides SDR qualification time and landing-page or form tooling instead of salaries and agency fees.
Ignoring the lag between spend and lead delivery
In longer-cycle B2B campaigns, this month's leads often trace back to last month's spend or content investment. Dividing this month's spend by this month's leads produces a noisy number that moves with timing, not performance.
Where we work on this
The decision this number feeds into