Stop Optimizing for Cheap Leads. Optimize for This Instead.

Cheap leads feel like winning. The cost per lead drops, the volume climbs, the dashboard turns green. Then three weeks later sales tells you none of them close. This is the quietest way a lead-gen account fails, because the metric everyone watches keeps improving while the business gets worse.

Why optimising for CPL backfires

When you tell Google or Meta to get you the cheapest possible lead, it does exactly that. It finds the people most willing to fill in a form, which is not the same as the people most likely to buy. The platform optimises toward the click-happy and the tyre-kickers because that's what 'cheap lead' rewards. Your CPL falls and your close rate falls with it.

Optimise for the qualified lead, not the lead

The fix is to change what the platform is optimising toward. Instead of counting every form fill as the conversion, we feed the platforms a signal for lead quality — so they learn to find more of the leads that actually turn into customers, not just the cheapest ones.

• Pass offline conversions back: when a lead becomes a qualified opportunity or a sale in your CRM, send that event back to Google (via offline conversion import) and Meta (via the Conversions API).
• Optimise campaigns toward the qualified-lead or sale event, not the raw form fill.
• Add friction on purpose: a slightly longer form or a qualifying question filters out the least serious leads before they ever count.

Measure cost per qualified lead and cost per sale

Once quality data is flowing, the metric that matters becomes visible: not cost per lead, but cost per qualified lead and ultimately cost per sale. It is completely normal for the right answer to be a higher CPL — because a ₹300 lead that closes at 20% beats a ₹80 lead that closes at 2% every single time.

Do this instead

Stop optimising for the cheapest form fill. Wire your CRM outcomes back into the ad platforms and optimise for qualified leads and sales. The CPL might go up. The revenue will go up more.