There is a quiet win that most people managing Google Ads accounts have enjoyed without ever really naming it.
You set a Target CPA of $60. The campaign is limited by budget. Over a few months, Smart Bidding settles it at $38. You did not ask for that. The system found the efficiency inside the budget constraint and handed it to you, and every month you reported a cost per lead well under target and everyone was pleased.
On 17 August 2026, that stops.
Google is changing how target-based bid strategies behave on campaigns that are limited by budget. After the change, those campaigns will optimise more consistently toward the target you actually typed in, rather than the better number the system happened to find.
Google’s own example is blunt: if your Target CPA is $10 and your recent actual CPA is $5, your campaign will deliver closer to $10 from 17 August 2026 if you make no changes.
That is nineteen days from the date this was published. This guide covers exactly who is affected, how to find your at-risk campaigns, what each of your four options actually costs you, and the one mistake we expect most advertisers to make.
Part 1: What is actually changing
The mechanic in plain terms
Today, when a campaign carries a “Limited by budget” status and uses a target-based bid strategy, two things happen that Google now considers problems.
The campaign can overperform against its target, delivering a materially better CPA or ROAS than you asked for. And performance can swing around unpredictably when you adjust budgets, which is why raising a budget on a well-performing budget-limited campaign so often made performance worse.
After 17 August, campaigns that are limited by budget and use a target-based bid strategy will more consistently perform toward the stated target, including when you adjust budgets.
Google frames this as predictability, and on its own terms that framing is fair. If you have ever raised a budget by 30% and watched your cost per lead jump 40%, this change is aimed squarely at that problem. Scaling becomes more linear.
The cost is that the free efficiency goes away.
Google’s own example, and one caveat
From the Help Centre: a budget-constrained campaign with a Target CPA of $10 that has been achieving a $5 CPA will start delivering closer to the $10 target once the update takes effect, unless you change something.
Worth noting, as Optmyzr has pointed out, that this is Google’s illustrative teaching example rather than a measured result from a real account. Nobody should read it as “your CPA will double.” What will happen is that the gap between your target and your actual performance narrows, and how much that costs you depends entirely on how wide that gap currently is.
If your target is $60 and you are delivering $55, this is a non-event. If your target is $60 and you are delivering $30, you have a serious problem to solve in the next nineteen days.
The Target ROAS version is worse, and almost nobody is explaining it
Every example Google publishes uses Target CPA, and that quietly hides the more alarming half of this change.
With Target CPA, overperforming means your actual cost is lower than your target. Being pulled toward target means you pay more per conversion, but you get more conversions for the same budget. There is at least a trade there.
With Target ROAS, overperforming means your actual return is higher than your target. Being pulled toward target means your return goes down.
Work through it. Your Target ROAS is 300%. The campaign has been delivering 500%. It is limited by budget, so your spend is effectively fixed at the cap. After 17 August, delivery trends toward 300%.
Return on ad spend is revenue divided by spend. If spend is capped and ROAS falls from 500% to 300%, revenue falls. That is not a trade between efficiency and volume. On a budget-capped campaign it is simply less money coming back.
Google’s answer is that you can now raise budgets predictably, which is genuine but requires you to spend more to stand still. If your budget is fixed because the business fixed it, that answer does not help you.
The direction of adjustment is opposite, and people will get this wrong
This is the part we expect to cause real damage in the next nineteen days.
To preserve current performance:
- Target CPA campaigns: lower the target. $60 target, $38 actual, move the target down toward $38.
- Target ROAS campaigns: raise the target. 300% target, 500% actual, move the target up toward 500%.
One goes down. The other goes up. They feel like opposite actions and they are the same action, which is aligning the stated target with reality.
One practitioner quoted by Optmyzr put the instruction exactly this way, recommending advertisers “start nudging your CPA targets down/ROAS targets up (slowly of course)” and watch how the change reacts.
Anyone working quickly through a long list of campaigns, applying the instinct that “efficiency means a smaller number,” will lower ROAS targets and make the problem considerably worse.
The rollout
The change begins on 17 August 2026 and rolls out over several weeks rather than switching on overnight.
Google’s FAQ adds a detail worth knowing: use caution with forecasts between 17 and 31 August. The forecasting tools will be working against shifting behaviour during that window, so anything they tell you in the second half of August should be treated as indicative at best.
Part 2: Who is affected
Campaign types included
| Campaign type | Affected |
|---|---|
| Search | Yes |
| Shopping | Yes |
| Performance Max | Yes |
| Demand Gen | Yes |
| Travel | Yes |
| Display | Already using the new behaviour |
| Hotel | Already using the new behaviour |
| App campaigns | No, keeps current behaviour |
| Video reach campaigns | No, keeps current behaviour |
| Video view campaigns | No, keeps current behaviour |
It applies across Google Ads, Search Ads 360, Display & Video 360, Google Ads Editor and the Google Ads API. PPC Land has covered the Display and Video 360 side separately.
Two things stand out. Display and Hotel already work this way, which tells you this is a consolidation of behaviour rather than an experiment. And Demand Gen is included, which matters given that Display campaigns are currently being migrated into Demand Gen. If you are migrating and adjusting targets in the same window, keep the two changes separate in your head or you will not be able to attribute anything.
For multi-channel campaigns like Performance Max and Demand Gen, Google also warns you may see shifts in how traffic is distributed across channels. That is a second-order effect worth watching, because a change in channel mix inside Performance Max can change lead quality even when the headline CPA looks fine.
The three conditions
A campaign is genuinely at risk only if all three are true:
- It uses Target CPA, Target ROAS, or Target CPC for Demand Gen
- It carries a “Limited by budget” status
- Its actual performance is materially better than the target entered in the account
Miss any one of those and this change does not meaningfully affect that campaign.
Google is explicit that campaigns which are not budget-constrained will not change behaviour as a result of this update. So the first move is not a sweeping account overhaul. It is a filter.
Who is getting notified
Google is sending account-level notifications to advertisers with any campaign that was limited by budget at any point in the last 12 months while using an affected strategy.
That is a deliberately broad trigger, and it has a side effect worth preparing for: accounts that have not been actively managed will surface campaigns people forgot were running. If you get a notification listing campaigns you do not recognise, that is not a bug. That is an audit finding.
Part 3: How to find your at-risk campaigns
Using the Bid Target Adjustment Tool
Google launched a Bid Target Adjustment Tool inside Google Ads on 6 July 2026. It surfaces historical campaign performance and lets you review and apply new targets before the enforcement date.
You reach it through the account-level notification. Inside, you can keep the existing target, apply a target based on the performance shown, or enter your own number.
Google will not adjust anything automatically. Nothing in this rollout changes your targets for you. If you do nothing, your targets stay exactly as they are and delivery moves toward them.
Doing it manually, which we would do anyway
The tool is a useful starting point and it is not a substitute for looking properly. For one or two accounts, the manual check takes about twenty minutes.
Step 1. Go to the Campaigns view and set the date range to the last 30 days.
Step 2. Add the Bid strategy type column and the Status column if they are not showing.
Step 3. Filter for campaigns with a “Limited by budget” status. Then filter bid strategy to Target CPA, Target ROAS, or Target CPC.
Step 4. For each remaining campaign, put the target and the actual side by side. Add a column showing the gap as a percentage. That gap is your exposure.
Step 5. Sort by spend. A 40% gap on a campaign spending $200 a month is an afterthought. The same gap on your largest campaign is the entire project.
Step 6. Before you act on any of it, check what the conversions actually are. Segment by Conversions > Conversion action and confirm you are looking at real leads or sales rather than phone taps, page views or form starts.
That last step is not optional and it is the one most people will skip. We found in our own accounts that a single misconfigured conversion action distorted a headline benchmark by 5x. If your campaign appears to be beating its target because it is counting soft actions, then “lock in that performance” is the worst possible advice, because you would be permanently anchoring your bidding to a number that was never real.
Fix the measurement first. Then set the target.
Part 4: Your four options, and what each one actually costs
Google offers four paths. Each is right in different circumstances and none of them is free.
Option 1: Keep your current target
What happens: delivery drifts toward the target you set. If you were at $30 against a $60 target, your cost per conversion rises toward $60. In exchange, you get more volume for the same budget, and scaling becomes more predictable.
When this is correct: when the target genuinely reflects what the business can afford, and you want volume more than efficiency. If $60 per lead is profitable and you have sales capacity you are not filling, taking more leads at $60 is a better business outcome than fewer at $30.
The trap: this is only correct if the target was set deliberately. In most accounts we audit, the target was set once, by somebody who has since left, based on a number nobody can reconstruct. Do not accept drift toward a target you cannot justify.
Option 2: Lower the target to match recent performance
What happens: you apply your actual performance as the new target. Performance stays close to where it is.
When this is correct: when current performance is genuinely good, verified against real conversions, and you would rather protect efficiency than chase volume.
The trap: you are hard-coding a number the algorithm found under a budget constraint. If you later raise the budget significantly, that target may no longer be achievable, and a target set below what a campaign can deliver chokes delivery rather than improving efficiency. Lock in $30 today and you may spend Q4 wondering why the campaign will not spend its budget.
Option 3: Set a custom target
What happens: you pick a number based on the business rather than on either extreme.
When this is correct: almost always, in our view. Google’s own example uses $7 as the middle ground between a $10 target and $5 actual performance, and that is closer to how the decision should be made.
Your target should reflect what the business can profitably afford, not what Google happened to produce. That means starting from margin and close rate rather than from the account. Our break-even ROAS calculator and ecommerce unit economics calculator will give you the ceiling, and for lead gen the lead quality cost calculator gets you to a true cost per qualified lead rather than a platform number.
The trap: none, other than doing the work. Which is why most accounts will not choose this option.
Option 4: Switch to Maximize Conversions or Maximize Conversion Value
What happens: the target constraint disappears entirely. The campaign spends its full budget chasing the most conversions or conversion value available.
When this is correct: when the campaign never had the conversion volume to justify a target in the first place. Google recommends evaluating Target CPA over 30 days including at least 30 conversions, and plenty of campaigns running targets are nowhere near that.
When we checked our own accounts we found seven of our ten Target CPA campaigns sitting below that threshold, several of them dramatically. For campaigns in that position, this change is a good reason to do something you should have done anyway.
The trap: Google is clear that because these strategies optimise to spend the full budget without a target, your actual CPA or ROAS will fluctuate as you adjust budgets. You are trading predictability for volume. That is a real trade, not a loophole.
If you are unsure which of the four applies, our Bid Strategy Selector walks the same decision.
Part 5: The option Google buries, and why it might be the best one
There is a fifth path in Google’s documentation and it gets less attention than it deserves: increase the budget.
Here is why it matters more than it sounds.
Before this change, raising the daily budget on a budget-limited campaign with an overperforming target usually caused fluctuation or a drop in efficiency. That is why so many advertisers learned to leave well-performing budget-limited campaigns alone. Touching the budget broke them.
After 17 August, campaigns optimise consistently toward the stated target regardless of the budget limit. Which means budget increases become predictable in a way they were not before.
Google’s own recommendations here are worth following:
- Keep a daily budget comfortably higher than your average daily spend, so the campaign is not restricted
- Check the Recommendations page to forecast additional conversions available at your current target
- After a budget increase, wait one to two conversion cycles before evaluating
The reframe is this: if you have been sitting at $30 against a $60 target because budget was the constraint, and $60 is genuinely profitable, then the correct response to this change is not defensive. It is to set a target you can defend and then fund it properly.
That is the version of this change where you come out ahead. It only works if you know your real numbers, which is the recurring theme of everything on this site.
If your campaigns are chronically budget-limited, our earlier piece on what limited by budget actually means and when it matters covers the diagnosis, and the budget calculator will tell you whether the number you have in mind is realistic for your market.
Part 6: The mistake we expect most advertisers to make
Opening the tool and clicking Apply on everything.
The Bid Target Adjustment Tool will show you recent performance and offer to set that as your new target. On a screen with forty campaigns and a deadline approaching, applying all of them takes one minute and feels like diligence.
What you have actually done is permanently encode, across your whole account, a set of numbers that were produced by an algorithm working inside budget constraints, measured on conversion actions nobody verified, in a period nobody chose.
Three specific ways that goes wrong:
You lock in a number built on soft conversions. If phone taps or page views are inflating your conversion count, your “actual CPA” is fictional and you have just made it official.
You lock in a seasonal artifact. If the tool’s lookback covers an unusually strong month, you are setting a target you cannot hold in a normal one. This is the same reasoning that makes us cautious about seasonality adjustments: short windows make poor evidence.
You cap your own growth. A target set at the efficiency ceiling of a constrained campaign becomes the thing that prevents you scaling it later.
Use the tool as a diagnostic. Make the decision yourself, campaign by campaign, starting with the ones that spend the most.
And do not change unaffected campaigns just because the update is coming. If a campaign is not budget-limited, this does not apply to it.
Part 7: The structural problem underneath this
Here is the part most coverage will miss.
Being “limited by budget” is often a symptom of structure rather than a genuine shortage of money.
Split an account into eight campaigns and you have eight separate daily budgets. Money sits idle in the campaigns with no demand today while the campaign that has demand hits its cap by mid-afternoon. Every one of those campaigns can carry a “Limited by budget” status while the account as a whole is not short of budget at all.
Consolidate the same account into two campaigns and the budget flows to wherever the demand is, automatically, every day.
When we ran this check across our own accounts, all 24 of our Search campaigns were below the conversion volume threshold that automated bidding needs, and the cause was fragmentation rather than account size. Fragmented accounts produce more budget-limited campaigns and more under-fed bid strategies at the same time, which means this August change hits fragmented accounts hardest.
So if the notification lands and you find fifteen budget-limited campaigns on targets, the honest question is not only “what target should each one have.” It is “should these be fifteen campaigns.”
Our Campaign Structure Recommender will give you a second opinion on that, and the 9-Pillar Scorecard covers the wider account if you want to check the foundations before making bidding decisions on top of them.
Part 8: What the PPC community is actually saying
This did not land quietly. The announcement generated significant pushback, and the argument is worth understanding because both sides have a point.
The objection
Kirk Williams of ZATO Marketing put the sharpest version of it on LinkedIn: why would Google build a system that stops chasing the most efficient auctions the moment a campaign hits its budget ceiling?
That is the question. If Smart Bidding is capable of finding a $38 cost per lead inside a constrained budget, deliberately switching that capability off looks less like predictability and more like Google reclaiming margin it was previously leaving on the table.
Joey Bidner, a freelance Google Ads manager and coach, published a post criticising the change that drew 71 reactions and 27 comments from paid search practitioners, with the broad sentiment being that campaigns which have quietly outperformed their targets, sometimes for years, will see costs rise regardless of anything the advertiser does.
Google’s response
Ginny Marvin, Google’s Ads Product Liaison, replied directly in that thread. Her position is that characterising the update as advice to “let the system spend more money” is not accurate, and that the change will not lead to spend increases.
Her supporting argument is the one in Google’s documentation: performance has often fluctuated unexpectedly in budget-limited campaigns when budgets change, which made it difficult to scale with confidence. The change is aimed at that.
She also confirmed to Aaron Levy on LinkedIn that account notifications were rolling out, pointing advertisers to the tool where they can review historical performance and apply target updates.
The counterpoint worth taking seriously
The most interesting contribution came from Mike Ryan, in the comments under Kirk Williams’ post, and it reframes the whole thing.
His argument is that Google is not making Smart Bidding less intelligent. It is correcting a system that had leaned too hard on safe, predictable auctions in budget-limited campaigns, rather than exploring the wider set of opportunities that still fall within an advertiser’s target.
Search Engine Journal described the same mechanic: Smart Bidding often enters only the auctions most likely to convert efficiently, producing stronger-than-expected CPA or ROAS, and Google says that was never the intended behaviour.
If that reading is right, then a budget-limited campaign delivering a $38 cost per lead against a $60 target was not being generous with you. It was being timid. It was skipping auctions it could have won profitably because it had learned to protect a number rather than pursue the goal you set.
Which reframes what happens after 17 August. You may not simply pay more per conversion. You may get meaningfully more conversions from the same budget, at a cost per conversion closer to what you said you were willing to pay.
Where we land
Both readings are plausible and nobody outside Google can prove which is right until the change rolls out.
Our practical position: the reading you should bet on is the one where your target is a real number.
If $60 genuinely reflects what a lead is worth to the business, the pessimistic reading costs you nothing meaningful and the optimistic one is upside. If $60 was inherited from somebody who left in 2023, both readings are bad news, and no amount of arguing about Google’s motives fixes that.
That is why Part 3 starts with verifying conversions and Part 4 sends you to margin and close rate rather than to the tool.
One tactical disagreement to know about
Google’s guidance suggests a clean, single target adjustment before 17 August is fine.
At least one practitioner running this in production is deliberately moving in smaller increments and watching each step before going further, expecting headaches for advertisers who are not prepared.
Both can be correct depending on the account. A campaign with high conversion volume and a small gap between target and actual can probably take one clean adjustment. A campaign with thin volume, a wide gap and a target you are not fully confident in is better moved in stages, with a week between changes so you can see what each one did.
We would default to a single considered change on high-volume campaigns and phased changes on everything else.
Part 9: What this signals
Step back from the mechanics and there is a pattern.
For years, the implicit deal with Smart Bidding was that you set a rough target, the system found efficiency inside your constraints, and you kept the surplus. Plenty of accounts were quietly subsidised by that surplus, and plenty of agency reporting was flattered by it.
That deal is ending. What Google is offering instead is predictability: set a target, get that target, scale the budget without the number moving.
That is a fair trade for advertisers who know their real economics. It is a worse deal for advertisers who set a target once, three years ago, based on nothing in particular.
The strategic consequence is that your target now has to be a real number. Not aspirational, not inherited, not a placeholder. Whatever you type into that box is increasingly what you will pay.
Which means the work moves upstream, to margin, close rate, lifetime value and what a customer is genuinely worth. The same conclusion we keep arriving at from different directions: conversion tracking and unit economics come before bidding, every time.
Part 10: A nineteen-day plan
Days 1 to 3: find the exposure. Filter for budget-limited campaigns on target-based strategies. Record target, actual, gap percentage and spend for each. Sort by spend. You now know the size of the problem.
Days 4 to 6: verify the numbers are real. Segment conversions by conversion action on every campaign in the list. Confirm you are looking at genuine leads or sales. Anything counting phone taps, page views or form starts is disqualified from this exercise until it is fixed.
Days 7 to 9: work out what the business can actually afford. Break-even ROAS from margins, or true cost per qualified lead from close rate. Not what Google produced. What the business can pay.
Days 10 to 13: decide campaign by campaign. Highest spend first. Keep the target, lower it, set a custom number, switch strategy, or raise the budget. Write down the reason for each decision, because in October somebody will ask.
Days 14 to 16: implement. Make the changes. Avoid sweeping simultaneous edits across the whole account, and leave anything you have just changed alone.
17 August onward: monitor, do not react. Expect fluctuation during the rollout. Treat forecasts between 17 and 31 August with caution, per Google’s own guidance. Give budget increases one to two conversion cycles before judging. Do not reverse a decision in week one.
Frequently asked questions
What is changing on 17 August 2026? Campaigns that are limited by budget and use a target-based bid strategy will optimise more consistently toward the target you set, rather than potentially overperforming it. It applies to Search, Shopping, Performance Max, Demand Gen and Travel. Display and Hotel already behave this way.
Will my CPA double? Only if the gap between your target and your actual performance is that large. Google’s $10 to $5 example is illustrative rather than measured. Your exposure is the gap between what you typed and what you are currently getting.
Which campaigns are affected? Only campaigns meeting all three conditions: a target-based bid strategy, a “Limited by budget” status, and actual performance materially better than the target. Campaigns that are not budget-constrained are unaffected.
Do I need to do anything? Not necessarily. Google will not change your targets automatically, so if your current targets accurately reflect your business goals, no action is required. Action is needed where your target no longer reflects what you actually want to pay.
What is the Bid Target Adjustment Tool? A tool that launched in Google Ads on 6 July 2026, reached through an account notification. It shows historical campaign performance and lets you apply updated targets. Treat it as a diagnostic rather than a set of instructions.
Should I just apply the tool’s suggested targets? No, not across the board. Applying recent performance as your target permanently encodes numbers produced under budget constraints, potentially measured on conversion actions nobody verified. Use it to see the data, then decide each campaign on its merits.
Why did I get a notification for campaigns I do not recognise? Notifications go to advertisers with any campaign that was limited by budget at any point in the last 12 months on an affected strategy. It surfaces campaigns that have not been reviewed in a while, which is worth treating as useful information.
I use Target ROAS. Do I raise or lower my target? Raise it. This is the opposite of the Target CPA answer and it is the mistake we expect to see most. If your target is 300% and you are achieving 500%, moving the target up to 500% is what preserves your current performance. Lowering it would make things considerably worse.
Why is the Target ROAS version worse than the Target CPA version? Because on a budget-capped campaign, ROAS falling from 500% to 300% at the same spend means less revenue coming back. With Target CPA you at least trade efficiency for more conversions within the same budget. With Target ROAS on a fixed budget, there is no equivalent consolation unless you raise the budget.
Is Google doing this to make more money? That is the accusation, and Google disputes it. Ginny Marvin has publicly rejected the framing that this amounts to telling advertisers to let the system spend more, and says the change will not lead to spend increases. The more sympathetic technical reading, argued by Mike Ryan, is that Smart Bidding had been over-indexing on safe auctions in budget-limited campaigns rather than exploring opportunities still inside the target. Nobody outside Google can settle it before the rollout.
Should I make one target change or several small ones? Google indicates a single clean adjustment before 17 August is fine. Some practitioners are moving in smaller steps and watching each one. We would make a single considered change on high-volume campaigns and phase it on anything with thin conversion volume or a wide gap between target and actual.
Is this a bad change for advertisers? It removes free efficiency, which is a real loss. It also makes budget increases predictable, which was previously a genuine obstacle to scaling. Whether you come out ahead depends on whether your target reflects what the business can actually afford.
What if my campaign never had enough conversions for a target anyway? Then switching to Maximise Conversions or Maximise Conversion Value is likely the right answer regardless of this change. Google recommends at least 30 conversions in 30 days before evaluating Target CPA, and campaigns well below that were never getting reliable target-based bidding.
Does this affect Performance Max? Yes. Google also notes that multi-channel campaigns like Performance Max and Demand Gen may see shifts in how traffic distributes across channels, so watch the channel mix as well as the headline number.
Can I still set a more efficient target than my current performance? Yes, at any time. Google notes your daily spend will likely be affected, and suggests modelling it with the bid simulator first.
When can I trust my forecasts again? Google advises caution with forecasts from 17 to 31 August. After the rollout settles, treat the first full month as your new baseline rather than comparing against pre-change periods.
About The Author
