Google Ads Bidding Change: Your Target CPA and ROAS Stop Being Suggestions on August 17

Starting August 17, 2026, Google Ads will stop letting budget-limited campaigns quietly beat their Target CPA or Target ROAS. If your campaign is marked "Limited by budget" and has been overperforming — actual CPA below target, actual ROAS above target — expect it to converge to the number you actually set. This is not a rumor: it is Google's own Help Center documentation, and it changes how you should audit, target, and scale paid campaigns from now on.

We manage Google Ads accounts for ecommerce and lead-gen clients, and this update hits the exact pattern we see most often: a client account with a campaign sitting at 4x ROAS against a 3x target, "Limited by budget" in the status column, everyone celebrating the overperformance. As of August 17, that campaign will start drifting toward 3x. The celebration was never real — it was a bidding system under-constrained by budget. Now the constraint gets removed, and the target becomes the ceiling.

What actually changes on August 17

Google's framing is simple: today, when a campaign has a "Limited by budget" status and uses a target-based strategy, it can overperform its target and fluctuate unpredictably when you touch the budget. After August 17, budget-limited campaigns using Target CPA, Target ROAS, or Target CPC (in Demand Gen) will "more consistently perform toward your bid target, including when you make budget adjustments."

Google's own example is the clearest way to explain it: a campaign with a $10 Target CPA that has actually been converting at $5 will deliver closer to a $10 actual CPA starting August 17. Same logic inverted for ROAS: a campaign with a 300% target delivering 500% will trend toward 300%.

Two design consequences matter for operators:

1. Budget and efficiency become separate levers. Before, raising budget on a constrained campaign could swing efficiency in ways nobody asked for. After, budget controls volume and the target controls efficiency, cleanly.
2. Multi-channel campaigns shift internally. Performance Max and Demand Gen may redistribute traffic across channels as the system optimizes more strictly to your target. Expect the channel mix to move even where the headline number stays flat.

Who is affected (and who is not)

Note the trap: most coverage of this update says "Target CPA and Target ROAS," but Target CPC is also in scope inside Demand Gen. If you run Demand Gen with a Target CPC, you are affected.

The gap in your account means one of two things

Here is the part that requires honesty. A campaign that is beating its target is either:

• Deliberate headroom. You intentionally set a loose target (low tROAS, high tCPA) so Smart Bidding could explore, find new audiences, and learn — the strategy Google Ads freelancer Joey Bidner described as "some of my best-performing accounts" on LinkedIn, calling this "one of the most self-serving Google-centric changes we've seen in years."
• A stale number. Nobody has touched the target since it was set 14 months ago, and the gap is an accident of budget constraint, not strategy.

The size of the gap does not tell you which one you have. The check is whether the target survives contact with your P&L: does the tCPA match what you can actually afford to pay for that conversion? Does the tROAS clear your real gross margin, not the margin your client deck claims? A target that survives that check was deliberate. A target that doesn't was fiction, and August 17 just forced you to fix it.

Google Ads Liaison Ginny Marvin pushed back on the "self-serving" reading: the change "only impacts budget-constrained campaigns using a target because this is already the bidding behavior when campaigns using a target aren't budget constrained… we're making the controls clearer; the target will more precisely control your ROI." Both things can be true: Google is making the system more predictable, and a working exploration strategy is collateral damage.

The three-question test before you touch anything

We are running every client account through this before the deadline:

1. Was the gap on purpose? If the target was set loose deliberately for exploration, you are losing a real lever on August 17. Plan for it: tighten creative testing cadence and audience signals so the system does not need as much headroom to learn.
2. Does the target survive margin math? Run the tCPA against your max allowable acquisition cost and the tROAS against gross margin minus COGS and fulfillment. If the target fails, it was never a target — it was a wish.
3. Can the campaign scale at this target? A 4x ROAS at a $50/day budget is meaningless if the target forces 4x at $500/day too. The whole point of the update is that scaling becomes predictable: you can now raise budget and trust the efficiency holds.

The pre-August-17 checklist we are using

1. Export the last 30–60 days of actual CPA and ROAS vs. target for every campaign with a "Limited by budget" status.
2. Flag every campaign where actual beats target by more than ~15%. Those are the ones that will visibly move on the 17th.
3. Decide per campaign, not per account: keep the target (only if it survives the margin test), adjust to recent performance, set a custom number that reflects your real goal, or switch to Maximize Conversions / Maximize Conversion Value if you want volume-first with no efficiency promise.
4. Do not blind-click "Apply" in the Bid Target Adjustment Tool. The tool — live since July 6 — calculates a recommended target from recent performance and applies it in one click. That recommendation is often right and sometimes dangerous: matching recent performance locks in a number that may be too aggressive once you scale budget. Google does not recommend targets for campaigns with fewer than roughly seven conversions, so those still need a human decision.
5. Give the budget room before the change lands. Google's own guidance: keep daily budget comfortably above average daily spend, use the Recommendations page to forecast conversions at your target, and wait one to two conversion cycles after any budget increase before judging the campaign.

The honest take

This update kills fake good news, and that is the part agencies should say out loud. If you have been reporting 500% ROAS against a 300% target, you were reporting a budget artifact, not marketing performance. The Aug 17 change forces target hygiene that most accounts needed anyway.

But the exploration argument is real for a minority of accounts. If your strategy depended on loose targets to let Smart Bidding wander, the new world punishes it — the fix is to move the exploration energy into creative volume and signal quality, not to fight the bidding change. For a deeper look at how AI-driven bidding and programmatic decisioning are evolving, see our analysis of AI predictive bidding and programmatic buying, and for the measurement side of the same coin, what the engage-through attribution change did to Meta reporting.

If you have not looked at your Google Ads account since the notification arrived, start with the export in step one. The deadline does not move, and Google will not adjust a single target for you. If you want a second pair of eyes on your account before the 17th — including a full audit of budget-limited campaigns, target-vs-margin math, and a scaling plan — Mintec runs these audits for ecommerce and B2B accounts in both English and Spanish. You can also start with the basics in what Google Ads is and why companies use it daily, then compare how ROAS behaves across channels in our LinkedIn Ads ROAS breakdown.

The campaigns that win after August 17 will be the ones whose targets survived a margin check this week — not the ones that looked best on a dashboard.