Google Ads' Target CPA reset: what the August 17 bidding change means

For a while, a budget-limited Target CPA campaign could look like the best performer in the account. Set the target at $10, watch the dashboard settle at $5, and take the win. Google Ads has spent the past several weeks correcting that behavior, and the correction is not cosmetic: campaigns that were quietly outperforming their own targets are now being pushed back toward the number an advertiser actually typed in. Anyone who runs paid alongside organic and content work, which is most of the accounts we touch, needs to understand this before next month's board deck gets built on a number that is about to move.

What changed on August 17

Google's own framing, echoed by multiple advertiser-facing outlets tracking the rollout, is that budget-limited campaigns using a target-based bid strategy will now "more consistently perform toward your bid target," including when the budget itself gets adjusted. Before the change, raising a budget on a constrained campaign could cause performance to drift in ways nobody had asked for, because budget and target efficiency were tangled together under the hood. After the change, they are separated: budget controls how much you spend, and the target controls what you pay per result. Those two dials no longer fight each other.

The clearest description of the mechanism came from Google itself, in the example advertiser-facing coverage has repeated verbatim: if a campaign's Target CPA is $10 but its recent actual CPA has been running at $5, the campaign will deliver more closely to a $10 actual CPA once the change takes effect. That $5 number was never the deal. It was Smart Bidding finding cheap conversions inside a constrained budget and calling it efficiency, and the whole point of a target is that Google is supposed to hold it, not beat it by half.

Why Google made the change

Google's stated reasoning is consistency, not cost-cutting: campaigns limited by budget and running a target-based strategy should behave predictably relative to the number an advertiser set, instead of drifting to whatever price point clears the budget fastest. Under the old behavior, a below-target CPA looked like a gift, but it usually meant the algorithm was harvesting the cheapest available conversions and leaving volume on the table, which is a different trade-off than the one most advertisers thought they were making when they set the target in the first place.

“If your campaigns are running and you have a target CPA of 10 and it's actually bringing in five, you might see a change for more predictable performance. The change is that that five-dollar acquisition cost you have is going to be moved up to ten.”

That framing, from Greg Finn of Cypress North, is the one to internalize before anyone on the team panics about a cost jump next month. A gap between target and actual performance was never proof of skill. It was proof the campaign had room Google was quietly using, and the room is now closed.

The bigger pattern here is worth naming, because it shows up well beyond Google Ads. Automated bidding systems, AI visibility dashboards, cold email sequencers, they all report a number, and the number is only as trustworthy as the mechanism generating it. A budget-limited campaign quietly beating its target looks identical, on a screen, to a campaign that is genuinely efficient. The difference only shows up when you ask what the constraint was doing, which is the same question worth asking of any metric that looks a little too good before you have checked what produced it.

Who the change actually hits

The rollout applies to budget-limited campaigns only. A campaign that is not budget-limited, meaning it could spend more if better opportunities showed up, was already behaving close to its target and will not feel much movement. The accounts that notice are the ones running lean budgets against aggressive targets, which describes a large share of mid-market Google Ads accounts, especially ones layering paid capture on top of an organic and content program the way most of our clients do. If your account has grown budget in fits and starts, chasing whatever number the dashboard showed last quarter, it is more likely than average to have this exact mismatch sitting in it right now, waiting to correct.

Performance Max accounts deserve a closer look than the rest of the list, because Performance Max campaigns are frequently budget-limited without an obvious flag telling the account manager so, and because the asset groups inside a single Performance Max campaign can be pulling toward very different actual CPAs even while the campaign-level target looks stable. A Performance Max campaign that has been quietly running at half its stated target is not one clean number moving. It is dozens of underlying auctions all correcting toward the same target at once, which is why Performance Max belongs at the top of the audit list, not the bottom.

The three-step audit

None of this is a reason to distrust Smart Bidding generally. It is a reason to stop treating a below-target CPA as a permanent feature of an account instead of a temporary condition of a specific budget constraint. The teams that get surprised this month are the ones who never separated Google Ads performance from the budget math sitting underneath it, and who reported the $5 number to a board without asking what was actually behind it. We covered the same failure mode on the cold email side, where a healthy-looking reply rate hid a much smaller number of genuine positive replies, in why dashboard metrics mislead. The pattern repeats across every channel: a headline number that looks efficient because nobody checked the mechanism producing it.

It is also worth saying plainly that this is not a reason to panic-abandon Target CPA or Target ROAS as strategies. Both remain the right choice for accounts with a real, defensible cost ceiling per conversion. The change simply removes a form of hidden upside that was never guaranteed to begin with. Agencies managing accounts on a client's behalf have the added job of explaining this proactively, before a client opens their own dashboard and sees a CPA that jumped without a matching drop in lead quality, and assumes the account manager broke something. A short note ahead of the rollout, with the target-vs-actual gap already documented, turns a scary-looking chart into an expected, explained line item instead of a fire drill nobody saw coming.

Set targets to the unit economics, not the dashboard

The advertisers who come out ahead here are the ones who treat this as a forcing function. A target CPA or target ROAS should reflect what a business can actually pay for a conversion, not whatever number a constrained budget happened to produce last quarter. That is the same discipline we apply when aligning paid and organic budget for B2B SaaS clients: paid spend works best as a lever pulled on purpose, not a number that drifts because nobody checked whether the campaign behind it was budget-limited. The same logic applies across channels, including how outbound programs decide where the next dollar goes, which we broke down in our outbound budget allocation framework.

Optmyzr's rollout writeup puts the recommended patience at one to two full conversion cycles before judging the new numbers, and that timeline matters more than it sounds. A gap between target and actual reveals that something is off between the two, not which one is wrong. A campaign running below target might have been leaving real volume on the table, in which case the correction is good news wearing an ugly cost number. Or the below-target number might have reflected genuine, sustainable unit economics that the target itself should be updated to match. Google's tooling will not tell you which situation you are in. Only your own margin math will, and that math should have been driving the target in the first place, not confirming it after the fact.

One more practical note: the rollout is staged over several weeks, not applied to every account at once, so two campaigns of the same type in the same account can be on different sides of the change for a while. Do not assume a campaign is done adjusting just because a sibling campaign already settled at its new number. Check the actual CPA trend on each campaign individually before deciding whether the correction has finished landing, and keep the target-vs-actual log from the first audit step open until every flagged campaign has stabilized.

See where you are cited today

A free snapshot audit of your rankings and AI citations before we ever talk.