Google Clarifies the August Smart Bidding Update: What Actually Changes

What Google Announced - and Why Advertisers Pushed Back

Starting August 17, 2026, Google is changing how Smart Bidding behaves in budget-limited campaigns using Target CPA and Target ROAS. Today, a budget-limited campaign often quietly beats its target - delivering, say, a $35 CPA against a $50 target - because the system spends its constrained budget on the cheapest conversions available. After the update, Smart Bidding will optimize much closer to the target you actually set. Advertisers noticed, and the pushback was immediate.

The Behavior That Is Going Away

If your campaign is budget-limited, that gap between your target and your delivered performance is not a bonus Google owes you - it is a side effect of the budget cap. The system never had enough budget to chase the marginal conversions near your target, so it stayed in the cheap end of the pool. Come August 17, a $50 Target CPA will start behaving like a $50 target: performance closer to the number you set, in exchange for more volume within the same budget.

The Backlash - and the Signal Inside It

The immediate industry reaction was skepticism: why would the bidding system stop trying to be as efficient as possible? Some read the change more cynically - as Google engineering a reason for advertisers to spend more, or accept lower-quality traffic at the same spend. When a platform announces your campaigns will stop overperforming, that is a reasonable first response. But it misses the actionable message buried in the announcement.

"Google always whispers to those who listen," says AdStack's Matthew Hanni. "This is Google telling us, plainly: move your targets closer to your actual efficiency, or risk losing that efficiency. If your targets tell Google you are willing to pay a higher CPA after August 17, that is likely exactly what you will get."

Read that way, the update is less a rug-pull than a warning label. Google is telegraphing that your configured targets are about to be taken literally - and giving you six weeks to make sure they say what you actually mean.

Google's Clarifications

Google responded directly to the concerns. The key clarifications:

• Budgets will not automatically increase. Your caps stay exactly where you set them.
• Bidding targets will not change automatically. Google is not rewriting your tCPA or tROAS values.
• A new Bid Target Adjustment Tool will help you identify which campaigns are affected and adjust targets before rollout.
• The stated goal is predictable scaling - making delivered performance match configured targets - not increased spending.

One important caveat to "budgets will not increase": your spend still can. Many accounts are deliberately structured so campaigns are restricted by the target rather than the budget - the budget is set with generous headroom and the tCPA or tROAS acts as the real throttle on spend. In that structure, the update cuts both ways: once Smart Bidding optimizes to your target instead of beating it, it will buy the additional conversions available at that target, and spend rises toward the headroom in your budget. The result can be lower efficiency and higher spend at the same time - exactly the combination your target was supposed to prevent.

The Case That This Is Actually an Improvement

There is a credible argument the change fixes a real problem. Over the years, Smart Bidding in budget-limited campaigns has drifted toward playing it safe - harvesting the cheap conversions it already knows how to win rather than exploring for the incremental ones your budget could reach. And today's overperformance makes scaling unpredictable: double the budget on that "$35 CPA" campaign and the CPA can lurch anywhere, because $35 was never your real marginal cost - it was an artifact of the budget cap. When targets and delivery actually match, budget increases behave the way you expect them to. That predictability is worth something, provided your targets are set where you truly want them.

What to Do Before August 17

The pragmatic move is simple: any campaign currently beating its target should be evaluated, and in most cases the target should be moved closer to where the campaign is actually performing. After the rollout, Smart Bidding will treat your configured target as the goal - so if the target does not reflect the performance you depend on, you are handing that efficiency back.

• Inventory your budget-limited campaigns running Target CPA or Target ROAS, and flag every one currently beating its target.
• Check target-constrained campaigns too. If your budget has headroom and the target is what actually throttles spend, expect spend to climb toward that headroom as bidding optimizes to the target - make sure the budget cap is a number you would genuinely pay.
• Raise outperforming ROAS targets. If a campaign has a 400% Target ROAS but has been delivering 600%, evaluate raising the target toward 600% - otherwise the update will happily accept 400% and call it success.
• Lower outperforming CPA targets. Same logic, opposite direction: if the delivered $35 CPA is what your economics actually require, your target should say $35, not $50.
• Use the Bid Target Adjustment Tool when it appears in your account to confirm which campaigns are affected.
• Watch the first weeks closely. Volume should rise as efficiency moves toward target - verify the trade is one your margins can carry.

Your Targets Are About to Mean What They Say

The advertisers this update punishes are the ones whose targets never reflected their real economics - and the ones who will not notice until the September invoice. Knowing your true marginal CPA and ROAS takes conversion data you can trust. AdStack™'s PPC management and conversion attribution teams set targets from actual profit data, not platform defaults. Book a call before August 17 and we will run the audit with you.