I get asked at least once a month by a client or a prospective one: what's a good cost per lead. It's the wrong question, and I tell them that directly, because answering it with a number sends teams chasing the wrong outcome for months.
Cost per lead by itself tells you almost nothing about whether a lead-gen program is working. A form fill from someone who downloaded a whitepaper for a college assignment costs the same, on a CPL basis, as a demo request from someone with budget authority who's ready to buy this quarter. Treat those the same in your reporting and you'll optimize your way toward cheap, low-intent volume every time, because that's what a pure CPL target rewards.
Run the math on what the two leads are actually worth and the absurdity of averaging them gets specific. Say the whitepaper lead converts to revenue at half a percent and the demo request converts at fifteen percent against a $50,000 contract. The demo request is worth thirty times the whitepaper download, and CPL prices them identically. No other corner of the business tolerates accounting like that; nobody averages the value of a browsing visitor and a signed contract and calls the result a KPI. Lead generation does it every day, on the first slide of the deck.
I've run hundreds of lead-gen programs across B2C, B2B, advocacy, higher education, and trade associations, and the pattern holds across all of them. The programs that actually move revenue track cost per qualified lead, and further downstream, cost per sales-qualified opportunity, not raw CPL. Those numbers are harder to report cleanly in a monthly deck. They're also the only ones that tell you anything true about whether the budget is working.
How the cheap-lead spiral runs
Here's a real pattern I've watched play out more than once. A team is under pressure to lower CPL. They shift budget toward gated content and broad-match targeting, and CPL drops thirty percent in a month. Leadership is pleased. Three months later, sales is complaining that the pipeline is full of unqualified names that never respond to outreach, and the marketing team is defending a metric that looks great and means nothing.
What makes the spiral vicious is that the algorithms are in it too. Every platform's bidding system optimizes toward the conversion event you feed it, and if that event is any form fill, smart bidding becomes a machine for finding the cheapest humans on the internet willing to type an email address. It gets better at it every week. The account's reported efficiency climbs while its actual yield falls, and by the time the revenue gap is visible, two quarters of machine learning have been spent training the system to hunt exactly the wrong person. The algorithm did what it was told. CPL was the telling.
The fix isn't complicated, but it does require more discipline than watching one number. Define what a qualified lead actually looks like at your company, in writing, agreed with sales, before you launch a campaign. Track cost against that definition, not against every form submission. Feed the qualified event, not the raw one, back into the bidding platforms, so the machine hunts buyers instead of typists. And build in a feedback loop from sales back to marketing on lead quality that runs at least monthly, so the definition stays honest instead of drifting toward whatever makes the current numbers look good.
The lever after the form fill
What happens after the form fill matters just as much as the cost of getting there. I've seen strong CPL programs waste half their value because follow-up took four days instead of four hours, and speed to first contact is one of the most under-discussed levers in lead generation. The research on this has been consistent for over a decade: contact rates collapse within hours of a form submission, because the prospect who was comparing options on Tuesday afternoon has bought from someone else by Friday. A more expensive lead that gets a same-day response will outperform a cheap one that sits in a queue, almost every time, which means the routing rules in your CRM are quietly a media-efficiency decision worth six figures a year, owned by nobody in the media team.
If you're reporting CPL as your primary lead-gen KPI right now, I'd push you to add one more number to that same slide: cost per qualified lead, using a definition sales actually agrees with. Expect the first month to be uncomfortable, because the channel that wins on CPL and the channel that wins on CPQL are usually different channels, and the reallocation argument that follows is precisely the argument the program needed to have. That single change will tell you more about whether the program is working than a year of CPL trend lines ever will. It's the first thing we rebuild when a lead generation engagement starts at Echelon.