Google Ads has completed a global bidding change that can materially alter the performance of campaigns that are limited by budget and use target-based bidding.
The rollout began on August 17, 2026 and, according to Google, was completed globally on August 27.
The affected strategies include:
- Target CPA;
- Target ROAS;
- Target CPC for Demand Gen.
The key change is simple but important:
budget-constrained campaigns now optimize more consistently toward the target the advertiser actually entered.
That matters because some limited-by-budget campaigns historically performed more efficiently than their stated targets.
Those campaigns may now move closer to the configured target.
The old behavior could hide the real target
Google explains that budget-constrained campaigns could previously overperform their stated efficiency targets.
For example:
Target CPA: $10.
Historical actual CPA: $5.
Under the new behavior, Google says the campaign can move closer to the $10 target unless the advertiser changes the target.
That creates an operational issue.
A target that once functioned as a loose ceiling may now behave more like an actual optimization instruction.
Advertisers who were accustomed to “setting $10 and getting $5” need to reassess what the $10 setting means.
The change applies only to constrained campaigns
Google says Target CPA and Target ROAS campaigns that are not budget constrained are not affected by this change.
The update is focused on campaigns carrying a “Limited by budget” status.
It applies across major campaign types including:
- Search;
- Shopping;
- Performance Max;
- Demand Gen;
- Travel.
It also applies to relevant campaigns managed through Search Ads 360, and Demand Gen campaigns managed through Display & Video 360.
The auction itself is not changing.
Google describes this as a bidding-system change.
Why Google made the change
Google says previous behavior created unpredictability.
A campaign might overperform its target while constrained, then behave differently after a budget increase.
This made scaling less predictable.
The new model is intended to make budget adjustments more consistent.
If the campaign is configured for a $50 target CPA, Google wants the system to behave closer to that target whether the campaign is constrained or scaled.
That simplifies one kind of planning.
It also removes an accidental source of efficiency for some advertisers.
Independent analysis is already seeing consequences
Search Engine Land reported on September 17 that post-update analysis by Mike Ryan, Head of Ecommerce Insights at Smarter Ecommerce, suggested higher CPCs and lower impression share for some budget-limited campaigns after the change.
The interpretation in that analysis is that the old budget constraint could suppress bids.
When that suppression disappears, campaigns may bid more aggressively toward the configured efficiency target.
That can produce:
- higher CPC;
- different auction participation;
- different impression share;
- more volume at the stated efficiency target.
That is an external analysis, not a Google statement.
Advertisers should validate the effect in their own accounts.
The most exposed campaigns are the ones outperforming their target
Google specifically recommends reviewing campaigns that:
- are limited by budget;
- use target-based bidding;
- historically performed better than the configured target.
This is the key audit group.
Example:
Target ROAS: 300%.
Historical actual ROAS: 500%.
If the campaign was previously delivering 500% partly because of budget-constrained bidding behavior, leaving the target at 300% can allow the system to move toward 300%.
The advertiser may interpret that as performance deterioration.
From Google’s perspective, the system is doing what the advertiser asked.
The Bid Target Adjustment Tool is now live
Google has introduced a Bid Target Adjustment Tool to help advertisers review affected campaigns.
It can be accessed through account notifications or campaign bidding settings.
The tool is designed to help advertisers identify targets that may need to be updated to reflect actual business goals.
The tool does not change targets automatically.
Google explicitly says it will not automatically adjust:
- bid targets;
- daily budgets.
That responsibility remains with the advertiser.
Do not optimize from historical habit
The update forces a useful question:
Is the bid target the target the business actually wants?
In many accounts, targets accumulate historically.
A campaign may have a 400% ROAS target because:
- it worked last year;
- it was inherited;
- the campaign previously exceeded it anyway;
- nobody wanted to risk changing it.
The new behavior makes that less safe.
Targets need to reflect current economics.
Use unit economics to set the target
Target ROAS and Target CPA should connect to:
- gross margin;
- LTV;
- payback;
- conversion quality;
- cash constraints.
For ecommerce:
If gross margin is 25%, a 2x ROAS can still be economically weak.
For subscription:
A higher acquisition CPA may be acceptable if retention is strong.
For B2B:
Lead CPA may be irrelevant if downstream opportunity quality differs dramatically.
The advertising target should be derived from the business model.
Google recommends waiting 1–2 conversion cycles
If an advertiser changes targets after the rollout, Google recommends waiting one to two conversion cycles before evaluating performance.
That is especially important for long-conversion businesses.
A change made today may not be visible in final revenue metrics tomorrow.
Avoid reacting too quickly.
The update is fully live, and Google says forecasting models stabilized after the rollout completed on August 27.
What operators should do now
Build an affected-campaign view
Filter for:
- Limited by budget;
- Target CPA;
- Target ROAS;
- Target CPC in Demand Gen.
Compare target vs. recent actual
Look at:
- configured target;
- 30-day actual;
- pre-August behavior;
- post-August behavior.
Review CPC and impression share
Check whether:
- CPC increased;
- impression share changed;
- spend distribution changed.
Review multi-channel allocation
For Performance Max and Demand Gen, Google says traffic allocation across channels can shift.
Recalculate target economics
Do not simply restore the old actual CPA or ROAS.
Check whether the historical efficiency was necessary or accidental.
Avoid unrelated emergency changes
Google specifically advises against applying data exclusions or new bid limits solely because of this update.
The larger lesson
Automated bidding still requires human target design.
The algorithm can optimize efficiently toward the wrong business objective.
The August change makes that more visible.
Budget constraints should no longer be treated as a hidden optimization mechanism.
If an advertiser wants a $5 CPA, the target should increasingly say $5.
That sounds obvious.
In many accounts, it was not how the system behaved.
The operational priority now is to align configured targets with real economics before Smart Bidding does exactly what the settings tell it to do.
Build a before-and-after audit
Because the rollout completed in late August, advertisers now have enough distance to compare behavior around the change.
Create a campaign-level table with:
- target CPA or ROAS;
- actual CPA or ROAS;
- CPC;
- impression share;
- spend;
- conversions;
- conversion value;
- budget status.
Compare a pre-rollout window with a post-rollout window while controlling for major:
- promotions;
- seasonality;
- creative changes;
- conversion-tracking changes.
The objective is not to prove that every difference came from the bidding update.
It is to identify which campaigns changed enough to deserve investigation.
Do not confuse target alignment with profitability
A campaign can now deliver more predictably toward the configured target and still be economically wrong.
Example:
Target ROAS: 300%.
Business break-even ROAS: 450%.
The system can perform exactly as configured and destroy contribution margin.
That makes finance and bidding strategy more connected.
Operators should maintain a documented mapping between:
- platform target;
- gross-margin requirement;
- LTV assumptions;
- payback requirement.
The biggest risk after the update is not necessarily volatility.
It is leaving an outdated target untouched because historical overperformance used to compensate for a weak setting.
The new behavior makes configuration discipline more important.
