Meta Ads
What to do when your CPA doubles but your click-through rate stays the same
If your click-through rate held while your CPA doubled, the ad is still doing its job. People are stopping and clicking at the same rate. The money is leaking after the click, or the clicks have drifted from intent to curiosity. So leave the ad on, confirm sales are being tracked, walk the chain from click to purchase, and find what changed downstream.
When CPA doubles, the instinct is to go straight at the ad. New image, new hook, maybe a new audience. It’s the part of the account you stare at every day, so it gets the blame.
But a steady click-through rate is the ad telling you something. The same share of people are still stopping, reading and clicking. Whatever broke, it mostly broke after that.
So here’s the order I work through it in. Most of it is boring (the fixes usually are).
Step one: leave the ad alone for now
My rule for killing an ad is pretty strict. The reported CPA has to be three to four times my target, and the engagement has to be weak too, meaning click-through rate and cost per click. A high CPA with strong engagement usually means the ad is doing work I can’t see, so it stays on.
That matters here because Meta sequences your ads. Someone might see three or four of them before they buy, and depending on your settings, the credit usually lands on the first or the last one they saw. Turn off an ad people are still clicking and you can pull out a piece that was feeding your winner.
Two more checks before you panic. How much data is this? Under two full days is too early to judge anything. And what is the blended CPA across the whole account (total spend divided by total purchases), not just this one ad? A few bad days inside a healthy trend is a dip. A trend over a few weeks is a real problem.
Also, if you raised the budget right before this happened, read why your ads stopped working the day after a budget increase first. A steady CTR with a climbing CPA is exactly what paying for the next tier of buyer looks like.
Step two: make sure the sale is actually being counted
This one sounds too dumb to bother with. Do it anyway.
You’d be surprised how often the full funnel is just broken – it’s silly but very common. I’ve gotten on calls with people whose setup simply wasn’t working, and as soon as they fixed it, the sales started coming in.
A purchase event that stopped firing, a thank-you page that moved, a checkout that got updated and broke on mobile. Any one of those will double a reported CPA while your CTR sits right where it was. Go buy your own product on your phone and check that the purchase shows up where it should.
While you’re in there, compare what Ads Manager reports against what your payment processor says actually sold. The data inside Facebook is always a little skewed, which is why I keep my own count. If real sales didn’t drop, you don’t have a CPA problem. You have a reporting problem.
Step three: walk the chain after the click
Now read the funnel as a chain: clicks, then landing page views, then checkout, then purchase. Each link is a place the money can leak, and you’re looking for the one that got worse.
Start with clicks against landing page views. A link click counts when someone taps the ad. A landing page view only counts when the page actually loads for them afterwards (Meta needs your pixel on the page to see it). If clicks held and landing page views fell, the page is slow, the link is wrong, or something isn’t loading. That’s not an ad problem.
Then watch your cost per landing page view against your cost per click. When cost per landing page view climbs faster than CPC, the quality of the traffic is slipping even though click volume looks healthy. It’s one of the signals I use to tell a concept is wearing out.
If people are landing fine and just not buying, the leak is on the page or in the offer. Before you redesign anything, work out which page is actually losing the money in dollars, not percentages.
Step four: ask what changed downstream
Most of the time, when CPA jumps and CTR doesn’t, something changed after the ad. A new price, a new version of the sales page, a different checkout. Sometimes somebody else made the change and nobody mentioned it.
Someone who’d bought one of my trainings ran into this. He raised the price on his offer (which can be fine) but kept running the same ads with the same Post ID, and his results dropped off. My read is that the Post ID had built up its data on people buying at the old price, and when results slowed it started learning the wrong lesson. He changed his ads enough to get new Post IDs, and sales went up that Friday and more than doubled over the weekend.
The flip side happened to me on a low-ticket training I launched to cold traffic. For about three days I got two sales, while roughly half the people who clicked were making it to the checkout page. Lots of checkout visits, almost no buyers. Very odd.
The problem was three buy buttons at the top of the sales page. People landed, smashed the shiny button before they’d read anything, and bailed at checkout because they weren’t sold yet. I deleted those top three buttons so folks had to get through the video and a chunk of the page first. Slightly fewer people reached checkout, but 40% of the ones who did bought, and I settled in around a $40 CPA on ice cold traffic.
Same ads the whole time. The fix was on the page.
Step five: if the funnel is clean, look at who’s clicking
Sometimes the tracking is fine, nothing changed, and the page converts the way it always did for the right people. You’re just getting fewer of the right people.
When you have lots of clicks at a strong CTR and they aren’t converting, it’s almost always one of three things. Your setup’s wrong, there’s no real intent behind the click, or the offer isn’t right for the people clicking. Step two handled the first.
The intent one is sneaky, because a curious clicker and a serious buyer look identical in the CTR column. If you’ve shifted your messaging toward something more intriguing, you may be pulling people who want to know what the ad is about rather than people who want the thing. Whatever you ask Facebook for, that’s what you get more of. Go back through the language you’ve changed lately and ask whether it still speaks to someone who has the problem and wants it solved.
And the least fun possibility is the offer’s price in that market. I had a client with a high-ticket backend whose front-end ads dried up out of nowhere. He was doing all the “right” things, 40-50+ new ads a week and over $1,000 a day in spend, and his CPA tripled anyway. He’d run out of people willing to pay $197 up front. We moved his front end to $97 and he told me it was the most sales he’d ever had.
If fresh ads keep failing on a clean funnel, that’s an offer conversation, not a media buying one.
What to do this week
Take the two weeks before your CPA jumped and the days since, and put four numbers side by side for each: link clicks, landing page views, checkout visits, and purchases (from your processor, not just Ads Manager). Divide each one by the one before it. Whichever ratio fell is where this week’s work goes, and the ad stays on until you’ve found it.
Should I turn off a Facebook ad with a high CPA but a good CTR?
Usually not. I only kill an ad when its reported CPA is three to four times target and its CTR and cost per click are weak too. Strong engagement with a high CPA often means the ad is helping other ads close the sale. Judge the blended CPA across the account before you pull anything.
Why are my link clicks so much higher than my landing page views?
A link click counts the tap on the ad. A landing page view only counts when the page loads afterwards, which Meta measures through your pixel. A big gap usually points to a slow page, a broken or redirected link, or a pixel problem. Fix that before you touch the ad itself.
Why is my CTR high but nobody is buying?
It is almost always a broken setup, clicks with no real intent behind them, or an offer that doesn’t fit the people clicking. Check tracking first by buying your own product. Then check the page. Then look at whether your ads drifted toward curiosity instead of speaking to people who want the problem solved.
P.S. The fastest check on this whole list is buying your own product on your phone. It takes five minutes, and every so often it’s the entire answer.
