Your Target CPA Is About to Become Your Actual CPA

Starting August 17, 2026, Google Ads will make budget-limited campaigns on Target CPA or Target ROAS deliver at their stated targets instead of beating them. A campaign set to a $10 CPA that has been converting at $5 will drift toward the full $10. The Bid Target Adjustment Tool, live since July 6, gives advertisers six weeks to lower targets and lock in current efficiency before the shift begins.

Google announced the change on June 15 as part of a broader mid-year bidding and budgeting update, then followed up in early July with notification emails and a new help document explaining the move. The official framing says the update fixes "a longstanding inconsistency" where budget-limited campaigns could "overperform their bidding targets," and promises "more predictable scaling" and "less volatility when budgets move." I have read that paragraph several times now and I keep landing in the same place. Predictable, in this context, mostly means more expensive.

The mechanics, before the spin

The change applies to Search, Shopping, Performance Max, Demand Gen, and Travel campaigns that carry a "Limited by budget" status while running Target CPA or Target ROAS. Hotel and Display campaigns already behave the new way. App campaigns and the two video formats (reach and view) are excluded. And it lands everywhere at once: Google Ads, Search Ads 360, Display & Video 360, Ads Editor, and the API, per PPC Land's breakdown of the rollout.

The core mechanic is simple. Today, a budget-capped campaign with a $10 target CPA can quietly deliver conversions at $5 actual CPA, and the gap just sits there as extra efficiency you never asked for. After August 17, Smart Bidding will treat the number you typed into settings as the instruction, and delivery will move toward it over several weeks. Actual CPA (what you really paid per conversion) converges on target CPA (the number in the settings field). Same logic for ROAS, in the other direction.

The Bid Target Adjustment Tool, which went live July 6, surfaces every affected campaign and offers three outs: keep your target and accept the drift, apply a suggested target that matches recent actual performance, or type in a custom number. You can also sidestep the whole thing by switching to Maximize Conversions, or by raising budgets until the "Limited by budget" flag clears. What Google will not do is adjust anything for you. Six weeks of runway, then the system moves whether you opened the tool or not.

Why "predictability" reads like a price increase

Greg Finn of Cypress North was blunter than the help docs. He described the update as arriving "under the guise of predictability and optimization" and predicted "we are going to see increased CPCs," because, as he put it, "you have to see increased CPCs if you're saying our acquisition cost is going to double." That math is hard to argue with. The mechanism that lifts a $5 actual CPA to a $10 one is Smart Bidding paying more per click in the auction, and PPC Land's follow-up coverage walks through exactly that chain.

The cleanest way I can describe what has been happening until now: you placed a limit order at $10 and kept getting fills at $5. Nice while it lasted. The update rewrites the rules so that your limit price becomes the expected price, on the theory that you meant what you typed.

Google has been filling your order below your limit price. From August 17, your limit price is the price.

Barry Schwartz at Search Engine Roundtable pointed out how much supporting material Google shipped alongside this one: notification emails, a long explainer document, video content, and revised help pages. From what I've seen over the years, that volume of pre-emptive documentation usually tracks with how much pushback Google expects. Nobody publishes a defense brief for a change advertisers will love.

To be fair to Google, the old behavior was genuinely weird. Budget-limited campaigns overshooting their targets produced real volatility when budgets moved, and anyone who has raised a daily budget 30% and watched CPAs lurch around for two weeks has felt it. The inconsistency was real. It just happens that the fix transfers the surplus from your side of the ledger to theirs.

The audit: find every campaign coasting under its target

This is a 15-minute job per account, and it should happen this week, not the week of August 11.

First, filter your campaigns for "Limited by budget" status combined with Target CPA or Target ROAS bidding. Those are the only campaigns in scope. Second, for each one, put last-30-day actual CPA (or actual ROAS) next to the stated target. Third, apply a threshold: if actual CPA is running more than 20% below target, that campaign has meaningful exposure. Google's own example in the help documentation is a 100% gap, $5 actual against a $10 target, and campaigns like that will roughly double their acquisition cost if nobody touches them.

Then decide per campaign. Lowering the target to recent actuals locks in today's efficiency but also caps how hard the system will push for volume. Keeping the target means accepting higher CPAs in exchange for fuller delivery against your budget. A custom number in between splits the difference. Or you move the campaign to Maximize Conversions and let budget, rather than a target, do the constraining, which is the trade Google's own Target CPA documentation has always implied for budget-capped campaigns anyway.

One timing note that matters more than it looks: Smart Bidding starts optimizing toward a new target within minutes, but it takes one to two conversion cycles to actually settle there, and Google explicitly warns against making multiple target changes inside a single cycle. If your sales cycle runs two or three weeks, you functionally get one adjustment between now and August 17. Pick your number carefully, because you won't get a clean second swing. (Worth a calendar block this week, honestly.)

Riding it out vs. resetting: where I'd land

For most lead-gen accounts, I'd set the target to the last-30-day actual plus about 10% headroom. You keep nearly all of the efficiency you've been enjoying, and the small buffer gives Smart Bidding room to breathe when auction prices wobble. Pinning the target exactly at recent actuals seems to invite delivery problems, especially in accounts with thin conversion volume.

The exception is campaigns that are genuinely starving. If you're budget-capped and leaving demand on the table, letting delivery drift up toward your stated target buys you more conversions at a higher unit cost, and for some businesses that trade is fine. It comes down to whether your marginal conversion is worth your stated target rather than your recent actual, and honestly, a lot of teams have never priced that out because the old system never made them.

The other piece of this is the targets themselves. Plenty of those $10 targets were set two years ago, by someone who may not work there anymore, and nobody revisited them because the campaign kept beating the number. The update is forcing a pricing conversation that probably should have been happening quarterly all along.

My prediction: blended CPCs on affected budget-limited campaigns climb 15 to 30% within two conversion cycles of the rollout, for accounts that never open the tool. Google's worked example is a 100% CPA increase, so 15 to 30 on clicks feels conservative if anything. This lands in the same season as Google collapsing enhanced conversions into a single account-level toggle and finally exposing where PMax spends your money. The pattern across all three: Google wants cleaner inputs and stated intentions it can optimize against, and it is done inferring what you meant.

Set-and-forget targets just became set-and-pay

I don't think this is a scandal, exactly. The old behavior was a quiet subsidy, and quiet subsidies end. But the advertisers who lose money on August 17 will lose it to a settings field nobody has looked at since the campaign launched, which stings more than losing it to an algorithm would. Fifteen minutes with the adjustment tool, sometime before mid-August, and the whole thing is a non-event. I'd just not bet on remembering in six weeks.

Notice Me Senpai Editorial