TLDR: A paid campaign we ran for our own agency had the strongest click-through rate and the lowest cost per lead of anything we had tried, and we shut it down anyway, because the leads it produced almost never turned into clients we wanted. The campaign was optimising for attention, not for fit, and the market analysis that proved it changed how we judge every campaign we run since, for ourselves and for clients.
The numbers on the dashboard were the kind you screenshot for a team update. Click-through rate well above our account average, cost per lead lower than any other campaign in the account, and a lead volume that kept our sales calendar full for six straight weeks. I was proud of it for most of that time.
What broke the story was going back through the sales calls those leads actually generated. Almost none of them turned into a signed client, and the few that did churned faster than clients from any other source we tracked. The ad had been built around a broad, punchy promise, and it was pulling in exactly the audience that promise attracts: people curious about the idea, not people ready to buy a service at our price point.
The Metric That Lied to Us
Click-through rate answers one question: did the ad make someone curious enough to click. It says nothing about whether that person was ever going to become a customer, and a campaign optimised hard enough for clicks will, almost by construction, start attracting people who click for reasons unrelated to buying.
Harvard Business Review's research on the value of keeping the right customers makes the case plainly: not every acquired customer is worth acquiring, and a business that measures acquisition success by volume alone will systematically favour the wrong ones. Our campaign was a clean example. It was excellent at the thing we were measuring and quietly damaging on the thing that actually mattered, which is the retained value of the clients it brought in.
The Market Analysis That Changed Our Mind
We ran a simple comparison across every active campaign: cost per lead against retained revenue per lead at the six-month mark, not at the point of signature. The picture reversed almost completely. Our highest cost-per-lead campaign, a narrower one built around a specific service and a specific client size, produced fewer leads but a retained revenue per lead several times higher than the campaign we had been proud of.
- The broad campaign: low cost per lead, high volume, low retained revenue per lead once churn was factored in.
- The narrow campaign: higher cost per lead, lower volume, and the clients who came through it were still clients six months later at a materially higher rate.
Michael Porter's five forces framework is usually applied to industry structure, but the underlying discipline, understanding where real, defensible value sits rather than where attention concentrates, is exactly what our two campaigns were testing against each other without us realising it. The broad campaign competed on attention, where every agency with a bigger budget wins eventually. The narrow one competed on fit, where we had an actual advantage.
What We Did Next
We killed the broad campaign the same week the comparison came back, which felt worse than it should have, because on the dashboard we were shutting down the thing that looked most successful. We reallocated the budget into three narrower campaigns, each built around one service and one client size, each with a headline that would only be interesting to someone who genuinely fit the profile.
A campaign that filters out the wrong lead before the click is worth more than one that filters nothing and asks the sales team to do it instead. The narrower campaigns produce roughly a third of the leads the old one did. The retained revenue from those leads, six months out, is higher in absolute terms, not just per lead, because the sales team is no longer spending most of its week on calls that were never going to close.
What to Check in Your Own Account
Pull cost per lead and click-through rate for every active campaign, then go back six months and pull retained revenue per lead from the same campaigns. If the campaign with the best top-of-funnel numbers is not also the campaign with the best retained numbers, you are very likely optimising for the wrong thing, and the fix is rarely more budget; it is a narrower promise that only the right person finds interesting enough to click. It is the same check a digital marketing agency should be running on its own paid accounts before it ever runs it for a client, and it is now built into how we review campaigns every quarter, ours included.