.png)
Google has updated how tCPA and tROAS behave in budget-constrained campaigns. Here’s what changed and what advertisers should review.
Request a free auditFor years, advertisers could run a Google Ads campaign with a target CPA of €20 and still see it consistently deliver conversions at €12. The same could happen with target ROAS: a campaign set to 300% might regularly achieve 450% or 500%. If the campaign was outperforming the target, there was little reason to question the setup.
That assumption now deserves another look. On August 17, 2026, Google began rolling out a change to target-based bidding for campaigns specifically marked “Limited by budget”, completing the rollout globally on August 27. The update applies to Target CPA and Target ROAS. This is not a blanket change to every campaign using those strategies. The campaigns to review first are those that are both target-based and budget-constrained.
Google says the goal is to make performance more consistent and predictable when budgets change. Previously, a budget-constrained campaign could materially outperform its stated target. Under the new behavior, Google says these campaigns will optimize more consistently toward the target that has actually been set.
Take a campaign with a €20 target CPA that has recently been delivering at €12. The new behavior may allow its average CPA to move closer to €20. That does not create additional demand or conversions. If spend stays unchanged and CPA rises, the campaign produces fewer conversions. The same logic applies to tROAS: if spend stays flat and ROAS falls, conversion value falls.
Targets are still average optimisation objectives, not fixed prices or guaranteed returns. Actual results depend on demand, auction competition, conversion rates, measurement quality and conversion delay. A €20 tCPA does not mean every conversion will cost €20, and a 300% tROAS does not mean every day will land exactly at 300%.
The update is easier to understand if we separate three different decisions:
Google presents the change as making budget scaling more predictable, but the extra volume does not come from the bidding update itself. It comes from a separate budget increase and the availability of additional demand.
The practical risk is a target that no longer represents the economics of the business. Imagine a campaign with a 300% target ROAS that has consistently delivered 500%. If the advertiser leaves the target unchanged and the campaign moves closer to 300% while spend stays flat, conversion value falls. Raising the tROAS toward recent performance may help protect more of that value. Increasing budget is a separate decision that may generate incremental conversion value if demand supports it.
The same applies to CPA. If a lead campaign has a €30 target but has been generating leads at €16, leaving the target unchanged means that, if average CPA rises while spend stays flat, the campaign generates fewer leads. Lowering the tCPA toward recent performance may help protect the efficiency and lead volume already being generated.
A recent travel account review at Equeco illustrates why this diagnostic matters. Several campaigns appeared to be underspending, but most were not actually budget-constrained. Their lost impression share due to budget was 0%, pointing instead to constraints such as bidding targets, rank or available demand. Very few campaigns in the dataset showed a genuine budget constraint.
Low spend therefore does not automatically mean this Google update is the explanation. Before attributing a CPA or ROAS movement to the new behavior, confirm that the campaign uses an affected target-based strategy and is genuinely marked “Limited by budget”.
Google recommends reviewing affected campaigns, particularly those that historically performed better than their targets. A practical audit is:
Performance can move for many reasons at the same time, and Smart Bidding needs enough conversion data to adapt. Google recommends waiting one to two conversion cycles before evaluating performance when conversion delays are longer. A few days of weaker CPA or ROAS are not enough to conclude that the update is responsible, so account for conversion lag and compare the result with changes in demand, auction pressure, tracking and other campaign settings.
The most useful outcome of this update may be that it forces advertisers to revisit what their targets actually mean. The Smart Bidding conversation should start with the economics behind those settings: what is a conversion worth, how much can we afford to pay for a lead or sale, what return do we need to remain profitable, and is there enough demand to justify increasing budget if we want incremental volume?
The distinction is simple. Keeping an outdated target can reduce the results generated from the same budget if efficiency deteriorates. Adjusting the target toward recent performance may help protect the efficiency and results already being generated. Increasing budget is a separate growth decision. The update itself does not create demand; it changes how closely a budget-constrained campaign may move toward the target you have set.
