Google Ads bidding update (Aug 2026): what it actually means for target ROAS / CPA

Starting 17 August 2026, Google is changing how target-based bidding behaves in budget-limited campaigns (tCPA / tROAS).

They’re calling it “more consistent and predictable performance around your targets, even when budgets change.”

A Bid Target Adjustment Tool also rolls out from 6 July 2026.

Let’s be clear about what this is actually about.

For years, target-based bidding has never really meant “you will hit the target exactly.”

If you set a 10x ROAS target, you might see:

• 5x in some periods
• 12x in others
• 20x when budget constraints + auction dynamics align

That variability isn’t a bug, it’s how Smart Bidding exploits auction-level opportunity.

What Google is changing here is not the existence of variability.

It’s the amount of upside deviation allowed when campaigns are budget constrained.

In budget-limited campaigns, Google is now explicitly moving toward:

tighter clustering of performance around your stated target

Which means something important:

If you’ve been consistently overperforming your target, that behaviour is now more likely to be pulled back toward the configured level unless you adjust it.

So no. This is not “10x means exactly 10x every time”.

That’s not how auction-based bidding works, and it’s not what Google is claiming.

This is about variance compression, not deterministic control.

Where this gets interesting is the condition that triggers it:

👉 campaigns that are limited by budget

In budget-limited campaigns, Smart Bidding performance can appear to fluctuate because the range of auctions it can participate in is constrained and can vary from day to day. For more context on budget-limited behaviour, see Your Google Ads Account Will Likely Have This Budget Inefficiency

This update tightens that gap.

The implication is subtle but real:

Your target is no longer just a directional signal.

It becomes a stronger anchor on realised performance when budget is the constraint.

From a practical standpoint:

• Budget still controls scale
• Targets still control efficiency
• But the system will now hold efficiency closer to the target than before

My take:

This isn’t a revolution in Smart Bidding.

It’s Google reducing volatility in constrained campaigns to make performance more “forecastable” and easier to scale.

Good for predictability.

Less room for accidental upside.

Same levers, slightly tighter system.