Google Ads Target-Based Bidding Changes August 17: Reset Targets or Pay More Per Conversion

The campaigns quietly beating their targets are about to stop

Somewhere in an ecommerce account there is a Shopping campaign that has sat in "Limited by budget" since spring. Someone typed a $40 target CPA into it a year ago, back when $40 was the number that made the margin work. It has been delivering conversions at $20 ever since, and nobody has touched it, because why would you touch a campaign that beats its goal. On August 17, 2026, that campaign starts drifting toward $40.

That is the change, and Google states it in one line on its own help page: "After August 17, 2026, campaigns that are limited by budget that use a target-based bid strategy will more consistently perform toward your bid target, including when you make budget adjustments so you can grow your campaigns with more predictable performance." A number that has been behaving like a ceiling starts behaving like a destination.

Google's own example teaches it faster than any explanation. "If your campaign's Target CPA is $10, but your recent actual CPA performance is $5, your campaign will deliver more closely to a $10 actual CPA starting August 17, 2026." Then the instruction, in Google's words: "Update your target to $5 to maintain your recent performance or to a target CPA of your choosing based on your business goals."

None of this is a surprise drop, and it should not be read as one. Google published the change ahead of the date, account notifications have been going out, and the tool built for the migration has been in Google Ads since July 6, 2026. The deadline is the story. What follows is what actually changes, what provably does not (your spend), which campaigns are in scope, and what is worth doing with the days that are left.

What Google is actually changing on August 17

The degree of freedom that is going away

A budget-limited campaign on a target-based bid strategy currently has a quiet degree of freedom. The budget binds before the target does. The system runs out of money before it runs out of room to bid, so it takes the cheapest conversions available and stops, which can land the campaign well under the target you typed. Google describes the status quo as campaigns that "may be overperforming on bidding targets and see performance fluctuations when budgets are adjusted."

That overperformance was never a promise. It was a side effect of the budget being the binding constraint, and it is the thing being removed.

Before August 17:

• The budget runs out before the bidding room does.
• The campaign takes the cheapest conversions it can find and stops.
• Delivered CPA or ROAS can sit well inside the target you typed.
• Raising the budget often scrambles that efficiency.

After August 17:

• The target is the number the system optimizes toward, capped budget or not.
• A previously overachieving campaign converges on the number you typed.
• Raising the budget is meant to buy volume at the same stated target.
• The budget cap itself is untouched.

The behavior that replaces it

After the change, the target is what the system optimizes toward regardless of whether the budget is capped. Google's framing is predictability: bids are "optimized more consistently to your target when your campaign has a limited budget," which is meant to make budget increases behave predictably instead of scrambling efficiency every time you raise a daily budget.

Read plainly, this is Google removing a discount that budget-constrained accounts have been collecting without asking for it. Whether that reads as a fix or a tax depends entirely on whether the target in your campaign is a real business number or a number someone guessed in 2025 and never revisited.

Google's worked example, and why it is the whole article

The $10 target delivering $5 is not a hypothetical Google chose to be dramatic. It is the shape of the problem. The gap between the target you typed and the performance you get is the exact amount of efficiency that is about to evaporate, and Google's remedy is a single edit: set the target to what the campaign actually delivers.

Google is explicit that it will not do this for you. "Google will not automatically adjust your bidding targets or budgets." The change is automatic; the correction is not.

Your budget cap does not move, and Google says so plainly

The spend question, answered on Google's own FAQ

This is the part where most secondhand coverage goes wrong, so it is worth being blunt. Google's FAQ asks whether the change will result in increased spend and answers: "No. This change will not directly result in increased spend for you. Your daily and monthly budget limits will always be respected."

Google Ads Liaison Ginny Marvin said the same thing publicly when advertisers started describing the change as a cost increase: "To be clear, this won't result in campaign spend changes," and her recommendation matched the help docs: "Our guidance for those with budget-constrained campaigns currently over-performing on their target is to ensure the targets are in line with your goals."

If you take one accuracy point from this post, take that one. Anyone telling you Google is about to raise your spend is describing a mechanism that does not exist. Your budget is still the cap, and it is still yours.

What rises instead

Cost per conversion. That is the honest version of the claim, and it is the one Google does not dispute. At a fixed daily budget, a campaign that stops beating its target buys fewer conversions with the same money, which is another way of saying each conversion costs more. On a Target ROAS campaign the same movement shows up as ROAS falling toward the stated target rather than sitting comfortably above it.

Google's FAQ concedes exactly this much and no more: the change "may impact campaigns that are overperforming on bidding targets if you take no action." Kept straight, the two columns look like this.

• Does not move: your daily budget, your monthly budget, the auction, the invoice ceiling.
• Can move: cost per conversion, ROAS, conversion volume at the same spend, and for Performance Max, how spend is split across channels.

The trade being offered is efficiency you were getting by accident, exchanged for scaling that behaves predictably when you raise the budget.

Why "Limited by budget" is the entire condition

If a campaign is not budget-constrained, nothing happens to it. Google's FAQ is unambiguous: "Target CPA and Target ROAS campaigns that are not budget-constrained will not change their behavior." Campaigns with room to spend already behave this way, in Google's words: "Campaigns with unconstrained budgets scale performance inline with the stated target and will continue to do so after this update." The change extends that behavior to the constrained ones.

So the audit filter is not "all campaigns." It is campaigns flagged Limited by budget that also run a target-based bid strategy. Everything else is noise.

Which campaigns are in scope, and which are not

The included list

Google's FAQ is explicit: "This update applies to Search, Shopping, Performance Max, Demand Gen, and Travel campaigns managed in Google Ads or Search Ads 360, as well as Demand Gen campaigns managed in Display & Video 360." The help page's eligibility section adds Display and Hotel to the affected set. Put together, the campaign types Google names are:

• Search
• Shopping
• Performance Max
• Demand Gen
• Travel
• Display and Hotel, which already run the new behavior

For an ecommerce advertiser, the two names that matter are Shopping and Performance Max, which is why this lands harder on retail accounts than on lead-gen ones.

The exclusions, and the two types already living in the new world

Two details get lost in most write-ups, and both come from Google's own eligibility note.

• Excluded: "App Campaigns, Video reach campaigns, and Video view campaigns (VVC) will continue using previous bidding behavior."
• Already migrated: "Hotel and Display campaigns already have the new bidding behavior."

The second one is genuinely useful context. If you run Display on a target strategy and it has behaved predictably against its target for a while, you have already seen what this feels like. Note also that if you run video reach or view campaigns, August 17 is not your date, and neither is it a reason to skip the audit on the Shopping side of the same account.

The bid strategies this touches

The strategies in scope are Target CPA and Target ROAS, plus Target CPC for Demand Gen specifically. Google's FAQ closes the obvious follow-up questions in one line: the change "does not impact manual CPC or Target Impression Share bid strategies." And for anyone bracing for a deeper platform shift, the same FAQ states, "This is specifically a bidding change. Google Ads auction mechanisms will not change."

Portfolio bidding and shared budgets are included, with the caveat that adjustments must be made at the portfolio or shared-budget level, and that "for constrained shared budgets, the impact is distributed uniformly across all campaigns in that group." Campaign total budgets are the exception: "Behavior for campaigns using campaign total budgets will not change."

What the drift looks like on a thin-margin catalog

A worked example at a fixed daily budget

Take the arithmetic at Google's own example gap, a target that is double the delivered performance, and hold the budget still.

A Shopping campaign spends $200 a day on a $40 Target CPA and has been delivering at $20.

• Daily budget: $200, before and after. It does not move.
• Today: $200 divided by a $20 delivered CPA is 10 conversions a day, roughly 300 a month.
• After August 17, with no edit: delivery converges toward the $40 you typed.
• Then: $200 divided by $40 is 5 conversions a day, roughly 150 a month.

Same invoice. Half the orders. That is what "no spend increase" and "higher cost per conversion" mean at the same time, and why both statements can be true without contradiction.

Run the same drift on a Target ROAS campaign and it reads as revenue instead of orders.

• Setup: $300 a day, stated Target ROAS 400%, actually returning 600%.
• Today: $300 at 600% is $1,800 of revenue a day.
• After convergence: $300 at the stated 400% is $1,200 a day.
• Difference: $600 a day of revenue, at identical spend.

The gap is not a billing line. It is the part of the return on ad spend you were quietly banking.

Where Performance Max adds a second variable

Performance Max makes this less legible, because the target is not the only thing moving. Google warns that "for multi-channel campaigns like Performance Max and Demand Gen, you may also see shifts in how traffic is distributed across different channels."

So a PMax campaign can converge on its target and reallocate spend across inventory at the same time, which means a week-over-week comparison in late August is measuring two changes at once. Google also flags a forecasting blind spot in the same window: "Please use caution when using forecasts from August 17 to August 31." Plan the read, not just the edit.

Who is most exposed

Three profiles carry the most risk, and they compound.

• Accounts with old targets nobody has revisited, where the number is a fossil rather than a decision.
• Thin-margin catalogs, where a target set with slack becomes unprofitable the moment the campaign actually spends up to it.
• Low-volume campaigns, because Google states plainly that "we don't calculate a recommended target for campaigns with fewer than 7 conversions," so the tool hands you nothing and the judgment stays yours.

What to do before August 17

Start from the filter, not from the campaign list

The audit is three moves, in this order.

1. Filter to campaigns whose status is Limited by budget and whose bid strategy is Target CPA or Target ROAS. Nothing else is in scope.
2. Sort by the gap between the target and the delivered figure. That gap is the exposure, campaign by campaign.
3. Decide per campaign. A gap near zero needs no action. A gap at Google's example size, roughly 2x, is where your August gets decided.

Google's notification net is deliberately wide, which is why an alert is not the same as a diagnosis: "These notifications are triggered for advertisers with any campaigns that were limited by budget in the last 12 months and use a target-based bid strategy impacted by the upcoming changes." A campaign that was constrained in November and is not constrained today can still generate an alert.

The five options Google gives you

Google documents five paths. They are genuinely different decisions, not five wordings of the same one.

Option	What it does	Best when
Keep the target	Performance drifts up toward it	The target is a real margin number
Match recent actual	Locks in today's CPA or ROAS	You want no change on August 17
Set a custom target	Lands between the two	Current efficiency exceeds what you need
Switch to Maximize	Spends the budget, no target	Budget is fixed and ROI can float
Raise the budget	Buys volume at the stated target	The target is right and demand exists

The Bid Target Adjustment Tool covers the two target-editing paths, matching recent actual performance and entering a custom number. It has been available since July 6, 2026, it shows historical performance per campaign, and applying its suggestion is one click. Google's own line on the safest path: aligning your target to actual performance before the date means "you should not experience noticeable volatility or changes to your existing performance."

Two traps worth avoiding

• Over-correcting with account-level levers. Google explicitly warns against it: "Taking actions like applying data exclusions or new bid limits solely in response to this update is not recommended, as it can cause performance fluctuations." Reach for the target field, not the blunt instruments.
• Reading results too fast. Google's guidance is to "wait 1 to 2 conversion cycles to evaluate the campaign's performance," and that discipline matters more than usual in a window where forecasts are already flagged as unreliable. Judging a bid change on three days of data is the same error as judging a creative on one day of CTR.

The efficiency you can no longer borrow from the auction

A quiet subsidy is ending

Strip the mechanics away and here is what happened. A lot of ecommerce accounts have been running on efficiency they did not earn and could not explain, produced by the budget binding before the target did. That subsidy is being withdrawn on a published date with a tool attached, which is about as fair as platform changes get.

The uncomfortable part is what it reveals. If your account only worked because a stale target was quietly under-spending its own ceiling, the account was not working. It was being carried.

The lever that did not move

Once the auction stops handing back accidental efficiency, the remaining ways to lower cost per conversion are the ones that were always yours.

• A better offer, which changes what a click is worth.
• A better landing page, which changes how many clicks survive.
• Tighter targeting, which changes who arrives.
• Creative that earns a cheaper click and a higher conversion rate before any bid strategy gets involved.

Bid math redistributes what the creative earns. It does not create it.

This is why creative supply keeps ending up on the same list as bidding changes. Google has spent the past year giving advertisers more ways to coordinate delivery, from video campaign groups to expanded automation, while the input those systems optimize over stays whatever your team managed to produce that month. When the platform stops handing back efficiency nobody paid for, the number of honestly different ecommerce video ads you can test stops being a nice-to-have and starts being the variable you control. If that testing spans more than one placement, each one gates uploads on its own numbers, which we collected in video ad specs by platform.

How Novoads solves the creative supply problem behind a tighter target

Novoads generates ad-ready UGC-style video from assets you already have. Upload a product image or write a script, pick an AI actor, and you get a vertical video you can download and run on any ad platform, including the Shopping and Performance Max campaigns you are about to re-target. A video runs roughly $2 to $11 depending on the model and length, which is what makes testing five angles a decision about time rather than about budget. The point is not one hero video. It is having enough genuinely different UGC-style ads in the account that a tighter target has something better to bid on. Whatever you use to produce that volume, know where the artwork came from: a copyright suit over an AI ad generator's meme templates is a live reminder that a subscription is access, not a rights clearance.

You can try it for $1, which covers 3 days of access and continues on the $49/mo Inicial plan. Cancel whenever you want.

A target you never checked is now a promise you have to keep

For years, the target field in a budget-limited Google Ads campaign was closer to a wish than an instruction. You typed the number you could afford, the budget ran out first, and the system handed back whatever efficiency it happened to find. On August 17 that field starts meaning what it says.

So the work before the date is small and unglamorous: filter to Limited by budget, compare target to actual, and decide, campaign by campaign, whether the number in the box is a business decision or a leftover. Google will not make that call for you, and it has said so in writing. The platform is done subsidizing targets nobody revisited. What it will still reward is an account that can put something worth bidding on into the auction, and that half of the job never belonged to the bid strategy anyway.