Google packed three bidding changes into one June announcement, and most of the coverage led with the wrong one. Promotion Mode got the headlines. The change that will move numbers in the most accounts is quieter: starting August 17, campaigns that are budget-limited while running Target CPA or Target ROAS will bid toward their stated targets instead of internally lowering bids to stretch the budget 1. In a capped campaign, that will surface as a volume drop that looks like a performance problem. It is not one, and the accounts that treat it like one will make it worse.
What changes on August 17
Today, when a Target CPA or Target ROAS campaign hits its budget ceiling, Google’s bidding quietly lowers bids to spread the budget across more auctions, buying volume at the cost of drifting from the target you set. The Google Ads August 17 bidding change ends that behavior for budget-limited campaigns: they will deliver more consistently toward the stated target instead 1. The change was announced June 15, 2026, it applies across Search, Shopping, Performance Max, Demand Gen, Travel, and Display, and account notifications began going out July 6 1.
Google’s own framing of the announcement is growth language, bidding and budgeting updates to scale with 2. Strip the framing and the mechanic is simple: the system stops discounting your target to buy you volume. Fewer conversions, priced closer to what you asked for.
The volume drop will get misdiagnosed
When a capped campaign stops shading bids down, spend concentrates in the auctions it can win at the stated target. Impressions fall. Conversions fall. Efficiency holds. Anyone reading the account week over week in September will see a performance dip that is, in fact, a policy change from July, and the reflexive fixes (new creative, restructures, target changes made in a panic) will muddy the read further.
A budget cap and a hard efficiency target were always a contradiction. Until now, Google resolved that contradiction silently, in favor of volume. From August 17 it resolves in favor of the target, which pushes the decision back where it belongs, with the operator. Three defensible answers: raise the budget, relax the target, or accept less volume at target efficiency. The only wrong answer is not deciding, because then Google’s new default decides for you and your reporting narrative never catches up.
We are not alone on the deadline. Common Thread Collective is telling brands to audit budget-limited campaigns before August 17 and to map Promotion Mode against their Q3 and Q4 promo calendars 3. When the ecommerce shops and the lead-gen shops agree on the same to-do item, it is probably real.
The other two changes riding in the same announcement
Promotion Mode is the headline feature: a beta for Search and Performance Max that lets you schedule a temporary ROAS-tolerance change plus extra daily budget for a peak window, with automatic reversion afterward 4. It productizes what disciplined operators already did by hand for every sale window: loosen efficiency, add budget, and, the step everyone forgets, put it all back afterward. The auto-reversion is the feature. Manual reversion is where peak-season accounts historically bled.
The same announcement expanded Smart Bidding Exploration to Performance Max campaigns without product feeds globally, plus a Shopping beta 4. We covered Exploration in May in our piece on what it does to search terms, and the mechanics have not changed; what changed is the surface area it now touches. Google claims campaigns using Exploration see, on average, an 18% increase in unique converting search query categories and a 19% increase in conversions 4. Common Thread Collective cites Google internal data putting Search campaigns using it at 27% more unique converting users on average 3. Every one of those numbers is Google’s own. Treat them as directional, not as a forecast for your account, and validate against your conversion data rather than the dashboard’s enthusiasm.
The naming reversal is noise, but readable noise
Separately, starting in June 2026, Google began unbundling “Maximize conversions with a Target CPA” back into a standalone Target CPA strategy, and “Maximize conversion value with a Target ROAS” back into standalone Target ROAS, with no change to bidding behavior 5. Microsoft went the opposite direction, consolidating targets into Maximize strategies in August 2025 5.
There is nothing to rebuild here, and anyone selling a restructure off the back of a rename is selling you their invoice. But the timing carries a signal. Google renaming strategies back toward explicit targets in the same summer it makes capped campaigns honor those targets reads like one editorial decision, not two coincidences. The target you type into the box is about to mean what it says.
The audit to run before the deadline
Filter every account for campaigns flagged Limited by budget, then note which of those run Target CPA or Target ROAS. That intersection is your exposure list, and building it takes about ten minutes per account. For each campaign on the list, make the call now, deliberately: raise the budget if the volume matters more than the target, relax the target if the cap is fixed, or accept the volume loss and write one sentence in the change log saying so, dated before August 17, so September’s reporting conversation takes thirty seconds instead of a meeting.
This kind of pre-deadline sweep is standing work inside our management engagements every time Google moves a default, and questions about whether that work fits your account are the kind our FAQ already answers. The mechanics are not hard. The discipline of doing it before the platform forces the issue is the whole job.
The change lands August 17. The campaigns that look broken in September will be the ones nobody looked at in July.