You open the account. Nothing is paused, nothing is disapproved, spend is where it always is. And your cost per lead is up forty percent from the spring.
There’s no error message for this. Platforms don’t tell you when the ground moves, so it reads like the market got harder, or like whoever runs your ads stopped trying.
Usually it’s one of four things, and three of them are fixable.
Your target was set when you knew less
If you use a target cost per acquisition, you picked that number at some point. Probably early, probably as a guess, probably before you knew what a lead was really worth to you.
Automated bidding optimizes toward the target you set. That sounds obvious until you notice the implication: a campaign that was quietly beating its target has room to drift back toward it, and that drift looks exactly like performance decay.
Google changed the behavior here in August 2026. Budget-limited campaigns using target-based bidding now optimize more consistently to their stated target rather than overperforming it. Google was explicit that it does not adjust your targets for you.
So if your campaign was limited by budget and outperforming a number you set two years ago, it may now be delivering the number you asked for, which is worse than what you had.
Check what target is actually set. Not what you remember setting.
Your competitors changed and you didn’t
Auctions are relative. If two competitors started bidding on your terms in June, your cost goes up without anything in your account changing.
This one is not fixable by tuning. It’s fixable by being worth more per click, which means a better offer or a better landing page, or by being more specific than they are so you compete for narrower terms.
Anyone who tells you they can fix competitive pressure with bid adjustments alone is selling you something.
You can tell this one apart from the others by checking impression share. If you are showing less often for the same terms at the same budget, somebody outbid you. If you are showing just as often and paying more per result, the cause is elsewhere on this list.
Want a second opinion on this?
Book a free consultationYou’re measuring a different thing than you were
This is the most common and least suspected.
If your form changed, if a phone number stopped being tracked, if someone added a “contact us” click as a conversion, then your lead count changed definition. Cost per lead is a ratio, and if the denominator quietly changed, the ratio moved without anything real happening.
Before you diagnose a spend problem, confirm you’re counting the same events you counted last quarter. If you can’t confirm that, that’s the actual finding, and it matters more than the number that prompted the question. The definitions are the place to start: what counts as a lead you have to pay for.
Your landing page got worse without changing
Pages decay. A form plugin updates and breaks on mobile. A phone number stops being clickable. Page speed drifts as someone adds tracking scripts. Nothing looks different to you, because you’re looking at the desktop version of a page you already know how to use.
Half of your traffic is on a phone, in a hurry, and has never seen the page before. Open it on your own phone, on cellular, and try to contact yourself. Most people who do this find something.
Do it on a phone you have not used to visit your own site recently. A cached, logged-in, autofilled version of your page loads faster and behaves better than the one a stranger gets, which is how a genuinely slow page can feel fine to the person who owns it.
And before you treat the move as real at all, confirm there was enough volume behind it to mean anything, which is the whole of how much ad spend before the numbers mean anything.
Work through them in this order, because it runs cheapest to most expensive.
Confirm your conversion tracking counts what it counted before. If it doesn’t, stop here and fix that first.
Look at your current bidding targets and when they were last changed. If a target is older than your knowledge of the business, it’s a guess you outgrew.
Open your landing page on a phone and try to become your own lead.
Then, and only then, look at auction pressure. It’s the one you can’t control, so it’s the last place to spend attention.
If your numbers moved and the explanation is “the market,” that’s worth a second look. Usually something specific changed, and specific things can be fixed.
Frequently asked questions
Is a rising cost per lead always bad?
No. If you raised your prices or moved upmarket, a more expensive lead that closes at a higher value is progress. The ratio only means something next to what a customer is worth.
Should I just lower my target CPA?
Careful. Lowering the target usually lowers volume too, because you’re telling the system to bid only where it’s confident. That can turn a cost problem into a no-leads problem.
How long before I can tell if a change worked?
Give it one or two conversion cycles before judging. Reading results after three days mostly measures noise.
Could it just be seasonality?
Often, yes. Compare to the same month last year rather than to last month, if you have the history.