Your cost per lead is falling and your revenue is not. Here is why.

Cost per lead is the most quoted number in performance marketing and one of the least useful in isolation. It is trivially easy to halve: loosen the form, drop the qualifying questions, bid on broader terms, run a giveaway. The dashboard turns green and the pipeline does not move.

We see the same pattern almost every month. An account has improved CPL by forty per cent year on year, and closed-won revenue from paid is flat or down. Nobody is lying. The metric is simply measuring volume at the top of a funnel whose composition has changed underneath it.

The three ways CPL improves for the wrong reason

1. You changed what counts as a lead

A newsletter signup, a pricing-page download and a demo request are not the same event, but if all three fire the same conversion action, the cheapest one dominates the average. When the ad platform optimises toward that conversion action, it will find more of whatever is cheapest — which is almost always the least commercially serious behaviour.

Fix: separate conversion actions by commercial intent, and give the platform only the one you actually want more of.

2. You broadened the audience and the intent went with it

Broad match, Performance Max and advantage-style campaigns will reliably reduce CPL because they find people who will convert on a form for reasons unrelated to buying. This is not the platform misbehaving; it is the platform doing exactly what you asked.

Fix: feed the platform a value signal, not a count. If your CRM can return a qualified or closed-won value, send it back as an offline conversion. The bidding changes immediately.

3. Your offer got easier and less relevant

An ebook converts better than a consultation because it asks for less. It also selects for people who want a document, not a supplier. Lead magnets have a place, but a programme built entirely on low-commitment offers builds a list rather than a pipeline.

What to measure instead

None of this argues for ignoring CPL. It argues for never looking at it alone. The minimum set we hold an account to:

• Cost per qualified lead — using your sales team’s definition of qualified, not marketing’s.
• Lead-to-opportunity rate, by channel and campaign — the single fastest way to spot a source that is generating volume and nothing else.
• Pipeline created per thousand spent — the number a finance director will actually engage with.
• Time to first meaningful sales conversation — a leading indicator that moves weeks before revenue does.

If a metric can be improved by making your business worse, it is not a target. It is a diagnostic.

A ninety-minute exercise

Export the last two quarters of leads with their source, campaign, and current CRM stage. Group them by campaign and calculate two columns: cost per lead, and cost per lead that reached your first real sales stage. Sort by the second column.

In most accounts we audit, the ranking inverts. The campaign with the best CPL sits in the bottom third by cost per opportunity, and a campaign somebody nearly paused because it “looked expensive” is carrying the pipeline.

That single table usually pays for the analysis several times over, and it does not require any new tooling — only that someone joins the ad data to the CRM data once and looks at it honestly.