Let me paint you a picture. You've been running a Target CPA campaign at $10, feeling pretty smug because your actual CPA has been cruising at $5. You're basically getting a two-for-one deal, right? The algorithm is overdelivering, your boss thinks you're a genius, and life is good.
Well, starting August 17, 2026, that free lunch is over.
Google's latest update to target-based bid strategies is one of those changes that sounds boring until you realize it might blow up your campaign performance overnight. The short version: campaigns that are limited by budget will now actually hit the targets you set, instead of sometimes outperforming them. Revolutionary concept, I know.
The "Feature" That Was Actually a Bug
Here's what's been happening behind the scenes. When your campaign had that "Limited by budget" status and you were using Target CPA or Target ROAS, Google's system would sometimes overshoot your efficiency goals. Your $10 target might deliver at $5. Your 400% ROAS target might hit 600%.
Most advertisers treated this like a pleasant surprise. Free efficiency! But it was never documented, never guaranteed, and frankly, never consistent. As one practitioner noted
, campaigns limited by budget would just as often struggle to hit their targets as exceed them. The behavior was unpredictable, which is the opposite of what you want from a bidding system.Google's fix is almost comically straightforward: your target-based bid strategy will now try harder to actually hit your target. If you set a $10 CPA, expect a $10 CPA. If you set a 300% ROAS, that's what you'll get.
The panic I'm seeing on LinkedIn is mostly from people who got comfortable with the unpredictability working in their favor. But here's the thing: building your marketing strategy on undocumented system behavior is like building your house on a foundation of "it's probably fine."
Who Gets Hit and How Hard
Not every campaign is affected equally. According to Google's documentation, this change applies to Search, Shopping, Performance Max, Demand Gen, and Travel campaigns. App Campaigns and Video campaigns continue with the previous behavior. Display and Hotel campaigns already operate this way.
The campaigns most at risk are the ones where there's a significant gap between your stated target and your actual performance. If your Target CPA is $10 but you've been averaging $5, you're looking at a potential doubling of your cost per acquisition. That's not a rounding error.
For Performance Max and Demand Gen campaigns, there's an additional wrinkle: you may see shifts in how traffic distributes across channels. The system will optimize more consistently toward your target, which could mean different channel mixes than you're used to seeing.
The Real Problem This Exposes
Let's be honest about what's actually happening here. A lot of advertisers set targets based on vibes rather than math. They pick a number that sounds good, watch the algorithm do its thing, and adjust based on whatever results show up.
This update forces a more disciplined approach. Your target needs to reflect what you actually want, not what you hope the algorithm will beat.
Smart Bidding strategies work best when you give them accurate signals. Target CPA tells Google to optimize for a specific cost per action. Target ROAS tells it to optimize for a specific return on ad spend. When you set these targets arbitrarily, you're essentially giving the algorithm bad instructions and hoping it ignores them.
The irony is that Google making the system more predictable is being treated as a problem. We've been complaining for years about black-box algorithms and unpredictable performance. Now Google says "fine, we'll hit your targets more consistently" and everyone panics.
What You Should Actually Do
Google released a Bid Target Adjustment Tool back in July 2026 specifically to help with this transition. If you haven't used it yet, now would be an excellent time.
Start by auditing every campaign that's currently limited by budget and using a target-based strategy. Compare your stated target to your actual performance over the last 30, 60, and 90 days. If there's a significant gap, you have a decision to make.
Option one: adjust your target to match your recent actual performance. If you've been hitting a $5 CPA with a $10 target, change your target to $5. This maintains your current efficiency but tells the algorithm what you actually expect.
Option two: keep your current target and accept that performance will shift toward it. Maybe you set that $10 target for a reason. Maybe your business can absorb higher acquisition costs in exchange for more volume. That's a legitimate strategic choice.
Option three: increase your budget so the campaign is no longer limited. This removes the constraint that triggers the behavior change in the first place. Of course, this assumes you have budget to add, which isn't always the case.
The worst option is doing nothing and hoping it works out. That's not strategy; that's gambling.
The Bigger Picture
This update is part of a broader trend toward more predictable, more controllable automated bidding. Google has been slowly tightening the relationship between what advertisers ask for and what the system delivers.
For B2B marketers especially, this matters. Our sales cycles are long, our conversion values vary wildly, and our tolerance for acquisition cost volatility is lower than consumer brands. A system that actually hits our targets, rather than sometimes overshooting and sometimes undershooting, is genuinely useful.
The catch is that it requires us to know what our targets should be. And that means doing the math on customer lifetime value, acceptable payback periods, and realistic conversion rates. The algorithm can optimize toward a number, but it can't tell you what number to pick.
Marketing is like dating, remember? You don't propose on the first ad impression. But you also don't show up to the date without knowing what you're looking for. Google just made it harder to wing it.
