CPL
Cost per lead
What it costs to generate one lead.
Spend divided by leads generated. A narrower sibling of CPA used where the conversion is a form fill, a call or an enquiry rather than a sale. It measures the cost of getting a name and nothing about whether that name is worth having.
CPL = ad spend ÷ leads
Key takeaways
- CPL measures the cost of getting a name. It says nothing about whether the name is worth having.
- It is trivially easy to improve by making leads worse, which is why lead gen accounts drift.
- The ceiling is profit per customer multiplied by the rate at which leads close.
- Cost per qualified lead is the real metric, and it needs the client CRM.
- Native in-platform forms convert far better and qualify far worse than on-site forms.
The easiest metric in marketing to improve badly
Three changes cut cost per lead immediately, and all three tend to cut lead quality at the same time. Broaden the targeting. Remove qualifying fields from the form. Switch from an on-site form to a native lead form that submits inside the ad platform.
None of these is dishonest, and each has a legitimate use. The problem is that they all move the reported number in the direction everyone wants while moving the thing the number is meant to represent in the opposite direction. Cost per lead falls; cost per customer rises.
What makes it insidious is that nobody has to intend it. An agency optimising in good faith toward the metric it was given will find these levers, because they are the ones that work.
How a lead gen account drifts
The pattern is consistent enough to predict. Month one to three, CPL improves and the reports look good. Around month four the sales team starts saying the leads are poor, informally, to people who are not in the agency meetings. By month six that has hardened into a view, and the agency arrives at a renewal with a stack of reports showing sustained improvement in a metric nobody trusts any more.
The agency is not usually at fault for the drift so much as for not having built the feedback loop that would have caught it. Nothing in the ad platform can see what happened after the form, so nothing in the ad platform can tell you the leads got worse.
The argument is also unwinnable without data. A sales director who believes the leads are bad and an agency showing a falling CPL are not disagreeing about facts; they are looking at different halves of a process. Whoever brings a qualification rate to that meeting decides how it goes.
Closing the loop with the client CRM
The structural fix is to feed the qualification outcome back into the ad platform, so bidding optimises toward leads that turned into something rather than toward submissions. Google Ads calls it offline conversion import; Meta does it through the Conversions API with an offline event set.
Mechanically it means the lead gets an identifier at submission, the CRM records what happened to it, and that outcome is sent back attached to the identifier. It needs the client CRM to cooperate, which makes it a relationship question as much as a technical one.
The lag is the practical limit. A lead qualified three weeks after the click cannot inform this week bidding, and in long sales cycles the loop may be slower than the optimisation window. Where that is true, the usable proxy is an earlier signal: a completed discovery call, a demo attended, a quote requested.
Even without platform integration, a monthly qualification rate reported alongside CPL changes the conversation entirely. It does not optimise anything, and it means nobody is surprised.
Working out what a lead is allowed to cost
Start from profit per customer, which is average deal value multiplied by gross margin. Subtract the profit the business wants to keep from each sale, and the remainder is what acquiring a customer can cost.
Then multiply by the close rate. If leads become customers one time in eight, and the business can spend €400 acquiring a customer, the ceiling on a lead is €50. If the close rate is one in three, it is €133. The close rate moves the answer by more than anything an agency can do to the campaign.
Both inputs belong to the client and neither is usually known precisely. Asking for them at kickoff frequently reveals that nobody has calculated the close rate, which is more useful to discover in week one than in month six.
- Profit per customerAverage deal value × gross margin
- Max cost per customerProfit per customer, less the profit to keep
- Max cost per leadMax cost per customer × lead-to-customer close rate
Native forms and on-site forms are not the same channel
A native lead form inside Meta or Google prefills what the platform already knows and never leaves the app, so completion rates are far higher and cost per lead is far lower. A form on the client site requires a page load, a tap out of the feed, and manual entry.
The gap in qualification is roughly the mirror image. Friction filters, and removing it removes the filter. A person who typed their details on a site had to decide to; a person who tapped a prefilled form may have done it with minimal intent.
Neither is correct in general. Native forms suit high-volume offers with a fast follow-up, and they need the follow-up to be genuinely fast, because intent decays quickly. On-site forms suit considered purchases where qualification matters more than volume.
What is not defensible is comparing their CPLs as though they were the same thing, or switching to native forms and reporting the resulting drop as an optimisation.
What it means for an agency
A falling CPL is not automatically good news, and saying so before it falls is what makes you credible when it does. Pair it with a qualification rate from the client’s CRM from month one. Asking for CRM access after a quality complaint looks defensive.
Where it goes wrong
- Comparing CPL between an on-site form and a native lead form. The native form converts far better and qualifies far worse; they are not the same channel.
- Reporting CPL without lead volume. Halving CPL while halving leads is not an improvement.
- Counting a lead at form submission when the client counts it after a qualification call. A gap that quietly doubles the number you report.
In short
Cost per lead measures the price of a name and nothing about its value, which makes it the easiest metric in marketing to improve by making the outcome worse. Set the ceiling from the client close rate and margin, report volume and a qualification rate beside it, and get the CRM loop built early, because without it an agency can produce six months of improving reports and lose the account anyway.
Frequently asked questions
What is a good cost per lead?
Work it out rather than looking it up. Profit per customer multiplied by the rate at which leads become customers gives the ceiling. A business making €400 of profit per customer and closing one lead in eight can pay up to €50 per lead before it stops working. Industry benchmarks describe different deal values, margins and close rates, so they cannot answer this.
Why are my leads poor quality if CPL is improving?
Often because of what produced the improvement. Broader targeting, fewer form fields and native in-platform forms all reduce cost per lead and reduce the friction that was filtering out unserious enquiries. The way to tell is a qualification rate from the CRM, which is the only place that knows what happened after the form.
What is the difference between CPL and cost per qualified lead?
CPL counts every submission. Cost per qualified lead counts only the ones the sales team judged worth pursuing, so it is always higher and it is the one that relates to actual revenue. Getting it needs data from the client CRM, which is why it is worth asking for access at kickoff rather than when a quality complaint arrives.
Should we use native lead forms or send traffic to a landing page?
Native forms produce more leads at a lower cost and less qualified ones, because they remove the friction that was filtering. They suit high-volume offers with a genuinely fast follow-up, since intent decays quickly. Landing page forms suit considered purchases. What is not defensible is switching to native forms and reporting the CPL drop as an improvement.
Related terms
Sources
Last reviewed
Keplent keeps track of these for you.
It holds what each client agreed to, watches what their campaigns actually did, and tells you in plain sentences what needs you this week.
Join the waitlist