Google Killed Bid Suppression. Your Best Google Campaigns Were Partly an Accident, and It Ends Before Q4.

If you run Google Ads and any campaign in your account shows the status “Limited by budget” on a Target CPA or Target ROAS strategy, open that campaign list this week and compare delivered performance against stated target. The campaigns that have been quietly beating their targets for years are about to stop, and Google has told you plainly that it will not fix your settings for you.

That is the whole recommendation. The rest of this is why, and what the decision actually costs.

What changed, and when

On August 17, 2026, Google began rolling out a change to target-based bid strategies for campaigns in “Limited by budget” status. The global rollout completed August 27. Per Google’s own FAQ, the new bidding behavior is now fully live across all affected accounts.

Scope, from the same document:

• Applies to Target CPA and Target ROAS across Search, Shopping, Performance Max, Demand Gen and Travel campaigns, in Google Ads or Search Ads 360, plus Demand Gen in Display and Video 360
• Also applies to Target CPC, but only for Demand Gen
• Does not apply to manual CPC or Target Impression Share
• Does not change behavior for target-based campaigns that are not budget-constrained
• Applies to portfolio bidding and shared budgets, where any adjustment has to be made at the portfolio or shared budget level

And the line that matters most: Google will not automatically adjust your daily budgets or your campaign bid targets.

The mechanic underneath it

Before August 17, a budget constraint did something advertisers were never told to expect. When Smart Bidding saw a campaign hitting its daily cap, it bid more conservatively, effectively stretching the budget across cheaper auctions. The result was a population of campaigns that substantially outperformed the targets their operators had set. A campaign with a $10 target CPA might have been delivering $5.

Google’s example in its own FAQ is exactly that scenario, and Google’s stated consequence is exactly what you would expect: that campaign now moves toward the $10 you asked for, unless you change what you asked for.

Read that carefully. Your best-looking Google campaigns may have been returning what they returned partly because of a side effect of a budget cap, not because of your targeting, your creative, or your feed. That is uncomfortable, and it is worth knowing before you build a Q4 plan on top of those numbers.

What the auction data shows

Smarter Ecommerce published post-rollout analysis, covered by Search Engine Land on September 17. This is third-party panel analysis rather than a Google disclosure, so treat it as directional rather than as a rate card. It is the best read available a month in.

• Median CPCs on budget-limited tROAS campaigns rose 15.8% after August 17, from €0.38 to €0.44
• Median CPCs on campaigns that were never budget-limited fell 13% over the same window, to €0.33
• Before August 17, the median budget-limited campaign lost roughly 45% of impression share to rank and only about 4% to budget. Afterward, rank losses fell to roughly 30% and budget losses rose to roughly 33%
• Overall median impression share for budget-limited campaigns fell from 40% to 31%
• Before the change, more than half of the budget-limited campaigns analyzed were beating their ROAS targets. Among campaigns that were not budget-limited, only 30% overdelivered and 57% landed on target

The impression share split is the tell. Those campaigns are no longer stretching a budget through cheap clicks. They are bidding up, winning pricier auctions, and hitting the cap sooner.

There is a second-order effect that nobody is talking about enough. If constrained advertisers are no longer hoovering up the cheap end of the auction, that inventory moved somewhere. The 13% CPC decline on unconstrained campaigns is where it went. Your competitor with no budget cap is currently buying clicks cheaper than they were in July.

The math on your P&L

Illustrative, but built the way an operator should build it. Take a Shopping campaign with a $1,000 daily budget, consistently flagged Limited by budget, tROAS target set at 400%, historically delivering 600%.

Monthly spend: $30,000. Monthly revenue at 600%: $180,000.

Assume 40% contribution margin after COGS, fulfillment, and payment fees, before ad spend.

Before August 17: gross profit $72,000, less $30,000 ad spend, leaves $42,000 of contribution.

After, with the target untouched: the system bids toward 400%. Same $30,000 spend, revenue drifts toward $120,000. Gross profit $48,000, less $30,000, leaves $18,000 of contribution.

Same spend. Contribution down $24,000 a month, a 57% cut.

Here is why this will not trip any alarm you currently have set. Your breakeven ROAS at 40% contribution margin is 250%. At 400% you are still profitable. Nothing turns red. Your ROAS dashboard shows a campaign performing exactly at target, which is what you told it to do. The loss shows up only in the contribution line, a quarter later, when someone asks why margin compressed in a quarter where revenue held.

Now raise the target to 600%, what the campaign was actually delivering. It will no longer spend the full cap, because the efficiency you are demanding does not exist at that volume. Say it spends $20,000 and returns $120,000. Gross profit $48,000, less $20,000, leaves $28,000 of contribution, and you have $10,000 of monthly budget freed.

The exact spend level at the higher target depends on how deep your auction is, so do not treat $20,000 as a forecast. The ranking is what matters and the ranking is not in doubt. Re-targeting beats doing nothing by roughly $10,000 a month in this example, plus the optionality on the freed budget.

Be honest about the ceiling, though. None of the three options gets you back to $42,000. That number was subsidized by a system behavior that no longer exists. The job now is picking the least bad option, not restoring last quarter.

The per-campaign decision

Pull every campaign with status Limited by budget on tCPA or tROAS. Sort by the gap between delivered performance and stated target. Then sort each one into a bucket.

Bucket one: it was overdelivering and the economics support more spend. Raise the budget rather than the target. If the campaign was returning 600% against a 400% target and 400% still clears your contribution hurdle comfortably, the constraint was never the right lever. Remove it and let the campaign scale at a target you actually want. This is the only bucket where the change is genuinely good news.

Bucket two: it was overdelivering and you cannot or will not raise the budget. Raise the target to match what it was actually delivering over the trailing 30 to 60 days. Accept that spend and volume will fall. Redeploy the freed budget. This is most brands, most of the time, and it is the fastest money in this entire article.

Bucket three: low volume, fewer than seven conversions. Google will not calculate a recommended target for these, per its own FAQ. Nobody is going to do this for you. Set these manually against your margin, or consolidate them into a portfolio strategy so they have enough signal to bid with.

Bucket four: budget is genuinely fixed and you want maximum output inside it. Switch to Maximize conversions or Maximize conversion value. Google itself recommends this path for advertisers with inflexible budgets who can tolerate ROI fluctuation. You give up target control and you get spend certainty. For a cash-constrained brand heading into Q4 inventory buys, that trade is often correct.

Use the tool, ignore the default

Google shipped a Target Adjustment Tool for this. It sits behind the “Review your campaign targets” notification banner at the top of the account, or under Campaigns, then the settings icon on the campaign, then Bidding, then Review campaigns. In Search Ads 360 it is the “Review bid strategies” banner instead, and it will not appear in your Google Ads account at all.

Use it to see the recommended targets. Do not accept them without running each one through your own contribution margin. Google’s recommendation is built from your recent delivered performance, which is a reasonable starting point, but Google cannot see your COGS, your freight, your return rate, or the fact that half your conversions come from a SKU you are discontinuing in November.

Two settings notes worth having. Google says it takes one to two conversion cycles for performance to stabilize after a target change, so make the changes now and not in the second week of November when you cannot afford a learning period. And Google explicitly advises against applying data exclusions or new bid limits as a reaction to this update, because those cause their own volatility.

The other thing that happened this week

On September 16, Google began rolling out a Spend Benchmarks report inside Account Overview, showing whether you spend more, less, or about the same as businesses Google considers similar, on both spend and clicks. Google has not published how a peer set is constructed or how many accounts sit in one.

It is useful for exactly one thing: telling you whether a cost change you are seeing is specific to your account or market-wide. Given the CPC divergence above, that is a genuinely handy thing to be able to check right now.

It is not a target. The peer number cannot see your margin, your AOV, your conversion rate, or your inventory position. A report that lands in the same month as a bidding change that makes constrained campaigns more expensive, and that shows you a peer spending more than you, is not a coincidence you need to act on. Profitability sets your budget.

Where the freed budget goes

If bucket two applies across several campaigns, you will have real money to redeploy. The auction data points somewhere specific.

Campaigns that were never budget-limited got 13% cheaper. Those are where incremental dollars currently buy the most. Before you move anything into a new channel, check whether your own unconstrained campaigns have headroom, because that is the cheapest incremental click in your account right now and it required no new build.

Second, note the timing. Peak season auction pressure builds from mid-October and does not let up until January. Whatever CPC you are looking at today is the cheap version. A brand that enters Black Friday with stale targets on budget-limited campaigns is compounding two cost increases at once, and the second one is seasonal and unavoidable.

What to do this week

1. Filter every Google Ads campaign to status Limited by budget with a tCPA or tROAS strategy. Export it.
2. For each, pull trailing 30 and 60 day delivered CPA or ROAS against the stated target. Flag every campaign where delivered performance beats target by more than 20%.
3. Sort the flagged campaigns into the four buckets above. Decide by contribution margin, not by ROAS.
4. Open the Target Adjustment Tool, read Google’s recommendations, and override the ones that do not clear your own hurdle rate.
5. Handle the sub-seven-conversion campaigns manually or consolidate them into portfolio bidding.
6. Make the changes now, so the one to two conversion cycles of stabilization happen in September and October rather than during peak.
7. Check your unconstrained campaigns for headroom before you redeploy anywhere else.
8. Pull up Spend Benchmarks once, to confirm whether your CPC movement is yours or the market’s, then close it.

None of this is financial advice, and the margin figures above are illustrative. Run the decision on your own numbers.

The uncomfortable part of this change is not the 15.8%. It is that a chunk of Google performance that brands have been attributing to their own optimization was a system artifact, and it is being taken back four weeks before the highest-stakes quarter of the year. The brands that reprice their targets in September will spend Q4 buying volume at a margin they chose. The ones that do not will spend it buying volume at a margin Google chose for them.

If you want a second set of eyes on which of your campaigns fall into which bucket, or you want the retarget modeled against your real contribution margin before Q4 rather than after, book a call.