Eight placements, one budget, six months. The ranking by click-through rate and the ranking by cost per lead were almost exactly inverted. Here is the data and what it costs you to ignore it.
Over six months a single acquisition account produced 244,418 impressions, 5,026 clicks and 266 leads at a $20.36 blended cost per lead. That top line is unremarkable. The distribution underneath it is not.
Cost per lead by placement, cheapest first: Instagram Stories $13.58, Facebook Marketplace $16.95, Facebook Search $17.02, Instagram Feed $17.04, Instagram Reels $19.32, Facebook Stories $25.59, Facebook Reels $29.55, Facebook Feed $36.38.
Now the part that matters. Facebook Feed had the highest click-through rate in the account at 4.39%, and the worst cost per lead at $36.38. Facebook Marketplace had the lowest click-through rate at 0.91%, and was nearly the cheapest source of leads in the account.
A team optimising to click-through rate would have moved budget from the cheapest lead source to the most expensive one, and every dashboard would have shown the campaign improving while it got worse.
Click-through rate measures how well creative interrupts a scroll. Cost per lead measures whether the person who stopped was worth stopping. Those are different questions, and in a feed environment they are often in tension. Highly interruptive creative earns clicks from people with no intent, which is exactly what a cost-per-lead figure exposes and a click-through figure hides.
This is not an argument that click-through rate is useless. It is diagnostic: it tells you whether creative is working as creative. It is simply not a proxy for demand, and it should never be the number a budget decision is made on.
Splitting spend on cost per lead rather than engagement returned 69% more leads on identical budget. No new creative, no new audiences, no additional spend. The same money pointed at the placements that were already producing.
$20.36 is a real cost per registration. It is not a cost per qualified buyer. In this account roughly two thirds of the leads that were actually reached turned out to be renters rather than buyers, which puts true cost per qualified lead three to four times higher.
That gap is a creative and targeting problem, not a form-fields problem, and any agency quoting you a blended CPL without volunteering the qualification rate is quoting you half a number.
Report cost per lead by placement, not by campaign. Campaign-level averages hide exactly the inversion described above, because a single campaign can contain both your cheapest and your most expensive placement.
Then check the qualification rate on those leads before you scale anything. A placement that is cheap on registrations and terrible on qualified buyers is worse than an expensive placement that converts, and only a joined view of ad platform and CRM will tell you which is which.