The Update Everyone Read as Bad News
On August 17, 2026, Google changed how Smart Bidding behaves in campaigns that are limited by budget. Before the change, a campaign with a $10 Target CPA that was actually converting at $5 would keep converting at $5, because the budget cap was doing the rationing and the algorithm was quietly beating your number. After the change, that campaign delivers close to the $10 you typed in. Google's own help center uses exactly that example.
Most of the coverage treated this as a cost increase. "Your CPA is about to rise." "Google is ending target overperformance." Agencies rushed out audit checklists and told everyone to brace.
We read it the other way. This is the first time in years that a bid target has behaved like an instruction instead of a suggestion. If you set a target, you get the target. That is what Target CPA and Target ROAS were always supposed to mean, and it is what we wanted all along.
What Actually Changed
The scope is narrower than the headlines suggest. Three things matter:
- It only applies to budget-limited campaigns. If your campaign spends its full budget every day, the change touches it. If it does not hit the cap, nothing is different.
- It applies to target-based strategies. Target CPA and Target ROAS on Search, Shopping, Performance Max, Demand Gen, Travel, Display, and Hotel. Demand Gen's Target CPC is included. App campaigns and video reach and view campaigns keep the old behavior.
- Google does not touch your targets or budgets. The system now optimizes toward the number you set. If that number is stale, the results will look worse. If it reflects what you actually want, nothing bad happens.
Google also shipped a Bid Target Adjustment Tool. It lists your affected campaigns, shows recent actual CPA or ROAS against the set target, and lets you apply the recent performance as the new target in one click, individually or in bulk. It is under the Campaigns page, in the campaign's bidding settings, or from the "Review your campaign targets" banner at the top of the account.
Why We Think This Is a Good Change
The target was never supposed to be a ceiling
For years, an experienced buyer's trick was to set a loose target on a capped budget and let the algorithm find efficiency underneath it. It worked, but it worked by accident. You did not control the outcome, and the moment you raised the budget the efficiency evaporated, because the budget was the real lever and the target was decoration.
Now the target is the lever. You say 500% ROAS, the campaign pursues 500% ROAS. You say $40 per lead, it pursues $40 per lead. That is a control we have wanted since Smart Bidding replaced manual CPC.
Scaling stops breaking performance
This is the part the "CPA is about to rise" crowd skipped. Under the old behavior, raising the budget on an overperforming campaign was a coin flip. You would add 30% to the daily budget and watch ROAS fall off a cliff, because the algorithm had been hiding behind the cap. Google's help page now says the opposite outright: increase your budget and the campaign keeps optimizing to the stated target.
For an account that has been budget-limited at a strong ROAS for months, that is the single most valuable change Google has made to bidding in a long time. The ceiling on growth was never the auction. It was the fear of raising the budget.
Our Target ROAS accounts got better, not worse
We manage Target ROAS across a roster of e-commerce accounts, most of them spirits and DTC retail, and every one of them was reviewed before August 17 and re-targeted to its real goal. Since the change went live, all of them have shown significant improvement. Not a single account got worse.
The reason is simple. When we say 8x, we get something very close to 8x, and when we add budget the number holds. Before the update, the same accounts were at the mercy of whatever the cap happened to allow that week.
The Cost: This Is Now a Weekly Job
Here is the honest trade. Before August 17, a stale target was harmless. The campaign overperformed it, the budget did the work, and you could leave the target alone for a quarter.
Now a stale target is the whole problem. The campaign will do what the number says, so the number has to be right every week. Target CPA and Target ROAS management has gone from a set-and-review-quarterly task to a set-and-review-weekly task, or as often as your conversion volume allows.
That is more work. It is a lot more work across a roster of accounts. It is also work that produces a result you can actually steer, which is the point.
How often to update Target CPA and Target ROAS now
The cadence is a function of data, not the calendar. Our rule:
| Conversions in the last 30 days | Review cadence | Maximum move per change |
|---|---|---|
| Under 30 | Every 2 to 3 weeks | 10% |
| 30 to 100 | Weekly | 10 to 15% |
| 100 or more | Weekly, sometimes twice | Up to 20% |
Move the target in small steps and wait one to two conversion cycles before judging the result. Tighten a target by 40% in one move and you will throttle volume, then spend the next month recovering it.
The Playbook We Run
- List every campaign that has been limited by budget in the last 12 months on Target CPA, Target ROAS, or Demand Gen Target CPC. Google's notification only fires for those.
- Compare the set target to the trailing 30-day actual. Where actual is meaningfully better than the target, that gap is what the update takes away if you do nothing.
- Decide what the number should be. Not what the campaign got last month. What the business needs. A $5 CPA that was luck is not a $5 goal.
- Apply the new target through the Bid Target Adjustment Tool or the campaign settings. Bulk apply is fine for accounts with many capped campaigns.
- Raise the budget on your best campaigns. This is the new upside. Give them 20 to 30% more room and confirm the target holds over the next two cycles.
- Put the weekly review on the calendar and hold to the cadence above. This is the step most advertisers will skip, and it is the step that decides whether the update helps or hurts.
In our accounts the daily monitoring runs through Ad Campaign Concierge™, which flags any campaign whose trailing CPA or ROAS drifts more than 15% from its target so the weekly adjustment is a decision, not a hunt.
Who Should Worry
Two groups are actually exposed:
- Advertisers with dozens of capped campaigns and loose targets who have not opened the adjustment tool. Their performance has already shifted toward the stale numbers and will keep shifting until someone updates them.
- Anyone who cannot commit to the cadence. If nobody in the account will touch targets weekly, Maximize Conversions or Maximize Conversion Value without a target is the safer strategy. You lose control, but you also lose the risk of a wrong instruction being followed precisely.
Everyone else gets a better system. You set a number, you get the number, and you can scale without the floor falling out. That is what we asked for.
Frequently Asked Questions
Does the August 17 update affect campaigns that are not limited by budget? No. The change only applies to campaigns with a "Limited by budget" status. Uncapped campaigns behave as before.
Did Google change my targets automatically? No. Google does not adjust targets or budgets. The Bid Target Adjustment Tool suggests a target based on recent performance, but nothing changes until you apply it.
Does this affect Performance Max? Yes. Performance Max campaigns on Target CPA or Target ROAS with a budget limit are included. Google also notes that multi-channel campaigns like PMax and Demand Gen may see traffic shift between channels.
Should I switch to Maximize Conversions to avoid it? Only if you will not maintain targets. Maximize Conversions spends the full budget with no efficiency floor, so CPA and ROAS will fluctuate as budgets move. A maintained target is the better tool.
How much should I move a target at once? Ten to twenty percent depending on conversion volume, then wait one to two conversion cycles before the next change.