Google Ads Now Delivers Exactly What You Asked For
If you have opened Google Ads in the past few weeks, you saw the banner: "Review your campaign targets." It appeared in accounts everywhere, paired with a warning that campaigns with bid targets will behave differently after August 17, 2026.
That date has passed. The change is live in your account right now. Here is what it is, who it touches, and the decision it forces.
What changed on August 17
Google changed how bidding works for one specific group of campaigns: those that are limited by budget and use a target-based bid strategy. That means Target CPA, Target ROAS, and, for Demand Gen campaigns, Target CPC.
Before this change, a budget-limited campaign could quietly beat its target. You set a $10 Target CPA, the budget ran out before the algorithm ever needed to bid up to it, and you got conversions at $5. The budget was the real constraint. The target was a ceiling nobody touched.
That behavior is gone. Campaigns limited by budget now optimize consistently toward the target you entered, even when you adjust budgets. Google's own documentation uses a blunt example: a campaign with a $10 Target CPA that has recently delivered a $5 actual CPA will start delivering closer to $10 unless you change the target.
Read that again, because it is the whole update. The efficiency did not disappear because performance degraded. It disappears because you told the system $10 was acceptable, and the system now takes you at your word.
Who this touches
The change applies to Search, Shopping, Performance Max, Demand Gen, and Travel campaigns, whether managed in Google Ads or Search Ads 360, plus Demand Gen campaigns in Display & Video 360. Google will not automatically adjust your targets or your budgets, and the auction itself is not changing. Your daily and monthly budget limits are still respected. This is not a forced-spend change, whatever the louder commentary suggested.
You do not need to worry if...
Your campaigns are not limited by budget. Unconstrained campaigns already deliver in line with their targets and will continue to.
You run Maximize conversions or Maximize conversion value without a target attached. Those strategies spend the budget by design. Nothing changes.
You bid with Manual CPC, Maximize clicks, or Target Impression Share. All excluded.
You run App campaigns, Video reach, or Video view campaigns. Excluded. Hotel and Display campaigns already worked the new way.
Your actual CPA or ROAS already lands close to your stated target. If the target was a real number, this change is a non-event for you. That last group is smaller than it should be, which brings us to the point.
The target was never the problem. The guess was.
Here is the perspective the banner will not give you: what Google optimizes for is not necessarily what you want to optimize for.
Target CPA optimizes toward a cost per conversion. It has no opinion about whether that conversion is worth the cost. Target ROAS optimizes toward a revenue ratio. It has no opinion about your margin, your return rate, or whether that revenue includes customers you would have gotten anyway. Those are your numbers to know, and in a lot of accounts, nobody ever derived them. The target got set as a negotiation with the algorithm: high enough to get volume, adjusted when delivery stalled, nudged when a monthly report looked soft. It was a dial, not a decision.
The old bidding behavior forgave that. If your guess was too loose, the budget constraint saved you, and you banked efficiency you never asked for. As of August 17, the guess is enforced. The number in the settings is a promise in both directions: the system will pursue volume up to it, and it will stop protecting you below it.
What to actually do
The decision path is short, but each step is a real decision rather than a button.
First, find the exposure. Pull your budget-limited campaigns on Target CPA or Target ROAS and compare stated targets against actual delivered performance over the last 60 to 90 days. Google's Bid Target Adjustment Tool, live since early July, surfaces this from the banner itself, though it will not generate a recommended target for campaigns with fewer than 7 conversions.
Second, decide what the right number is. This is where I will push back on the tool's framing. Its headline option is "adjust your target in line with recent performance," one click, done. But recent performance at a constrained budget is not automatically the correct target at any budget. A $5 actual CPA might reflect the cheapest slice of available demand, the slice your limited budget happened to buy. Matching your target to it locks in efficiency but may cap volume harder than your economics require. The right target comes from your math: allowable CPA from margin and close rate, required ROAS from contribution margin and repeat purchase behavior. If $7 is what your unit economics support, $7 is the target, regardless of what last month delivered.
Third, pick the structure that fits the number. Lower the target to hold efficiency and accept the volume cap. Set a custom target and make the volume-versus-efficiency tradeoff explicitly. Or switch to Maximize conversions or Maximize conversion value if volume within a fixed budget matters more than a stable ratio, accepting that your actual CPA or ROAS will float as budgets move.
One genuinely good thing came with this. Raising the budget on an overperforming, budget-limited campaign used to invite volatility, which is exactly why many advertisers refused to scale winners. Under the new behavior, campaigns should hold their stated target as budgets grow. If your target is a real number and there is demand above your current cap, scaling is safer than it was in July. Be clear-eyed about the tradeoff: this update converts hidden efficiency into either delivered-at-target performance or additional spend, and Google benefits either way. That does not make the predictability claim false. It makes the accuracy of your target the entire ballgame.
The dividing line
This change will be invisible in some accounts and expensive in others, and the difference is not campaign structure or budget size. It is whether anyone ever did the math behind the target. Accounts where the CPA target traces to margin and close rates, where the ROAS floor reflects contribution rather than topline revenue, had real numbers in the settings all along. Nothing moves for them. Accounts where the target was a guess have been living on efficiency the algorithm found and nobody decided to keep. Google did not take that efficiency away this week. It just stopped delivering more than you asked for. Knowing what to ask for was always the measurement work behind the spend. Now it is enforced.
If you want the math behind your targets done properly, that is the work described on the Analytics and Reporting page, and the contact form is the fastest way to start the conversation.