Why your CPL went up (and why it's not Meta's fault)
Cost per lead doubled overnight. Your agency blamed iOS and the algorithm. Here's the real reason — and how to fix it in two weeks.
Every quarter, like clockwork, we get the same call. A founder of a $5M-$50M service business, frustrated, screen-sharing a Meta dashboard. CPL is up 40, sometimes 80 percent. The agency on the other end of the contract has a polished deck explaining that the algorithm changed, that iOS killed attribution, that Q4 is just hard.
Almost none of that is the actual reason. We've audited a few hundred of these accounts now. The pattern is so consistent it's almost boring.
The real reason CPL goes up
There are exactly four things that drive CPL on a service-business ad account, and only one of them is the algorithm. In order of how often we see it as the actual culprit:
- Creative fatigue — same five ads running for 90+ days
- Funnel rot — landing page hasn't been touched since launch
- Audience saturation — local market, same broad audience, no new creative angles
- Algorithm/auction shifts — real, but rarely the dominant factor
When CPL doubles, it's usually two of those at the same time. Stale creative driving a tired landing page is the #1 killer we see. The algorithm gets blamed because it's the only thing the agency can't fix and you can't argue with.
Creative fatigue is the silent killer
Meta's auction punishes ads with falling engagement. Once your hook-rate drops below ~25% and your hold-rate slides under 15%, you start losing impressions to fresher competitors at the same bid. The algorithm doesn't hate you — it's doing exactly what you'd want it to do if you were a competitor.
Most agencies refresh creative quarterly. That's catastrophic. By month three the winning ad from January is fatigued, the audience has seen it a dozen times, and you're paying premium for declining attention. We refresh weekly. Not because it's heroic — because the math forces you to.
If your top ad has been running for more than 6 weeks without a refresh, your CPL has already gone up. You just haven't noticed yet because the lag is slow.
Your landing page isn't helping
The second-biggest factor: your landing page hasn't been touched since the funnel launched. Headlines that worked in spring don't work in fall. Offers that converted at $3k AOV don't convert at $5k. The page that won 18 months ago is now an anchor.
We track Landing Page View → Lead conversion rate as the single most important diagnostic. If it's drifted from 8% to 4%, that's where your CPL doubled — not in the auction. The fix isn't a new ad. It's a new offer or a new hero section. Often both.
How to fix it in 14 days
If you've read this far, here's the playbook we run when we take over a stalled account:
- Day 1-3: Ship 4-6 new creative angles. Not new thumbnails — new hooks, new proof points, new founder talking-head cuts.
- Day 4-7: Rewrite the landing page hero, offer, and CTA. A/B test against current.
- Day 8-10: Audit attribution. Pixel + CAPI + GA4 + CRM events. Most accounts are silently miscounting.
- Day 11-14: Kill the bottom 30% of audiences and ad sets. Concentrate budget on what's lifting.
Done correctly, CPL drops 30-50% inside 14 days on most accounts we take over. Not because we're geniuses. Because the previous setup was bleeding from four wounds and only one of them was the algorithm.
The bottom line
CPL going up is information, not destiny. It's almost always a signal that creative, funnel, or both have gone stale and nobody's been minding the store. The algorithm gets blamed because it can't defend itself. Don't let your agency get away with that — make them tell you what they shipped this week.
If they don't have a real answer, you already know the problem.