Three Platforms Just Took Over Your Cost Controls
In six weeks, Google, Microsoft, and OpenAI each moved a piece of cost control out of your hands and into the platform's.
Three different mechanisms, one direction. Google made your cost target binding instead of a ceiling you could beat. Microsoft is removing the manual click cap from new campaigns. ChatGPT Ads changed which bidding option is already selected when you build an ad group.
None of these makes your ads more expensive on its own. The Google one does something less obvious and more annoying: it changes what you get for the same money without changing the invoice.
If you only run Google Ads, the first two sections are the ones that affect you and you can skim the rest. If you run Microsoft too, there is a date on this page you should not miss.
Prefer to have someone look at your accounts rather than read about them? Book a free strategy call. Otherwise, keep going — the whole thing is settings you already have.
Three Changes in Six Weeks
Worth separating, because they are not the same change and only one of them applies to everybody.
Google, August 17: Budget-Limited Campaigns Now Hit Their Target
This is the one that applies whether or not you ever set a manual bid, so it is the one to read.
Target CPA is the cost per acquisition you tell Google to aim for — in plain terms, what you are willing to pay for one lead. Until August 17, a campaign that ran out of budget before it could bid its way up to that number simply delivered under it, and you banked the difference.
It no longer banks the difference. The campaign moves toward the target you set, buying more expensive conversions with the same money and returning fewer of them. Google's own help documentation describes the update as delivering "more consistent and predictable performance for campaigns that are limited by budget." PPC Land walks through the arithmetic with a campaign delivering at $5 against a $10 target: under the new behavior it moves toward $10, total spend stays flat because the budget still caps it, and volume falls. That is one illustration, not a promise that everyone's costs double — how far your own targets sit above real delivery is the whole question, and it is different in every account.
It applies to Search, Shopping, Performance Max, Demand Gen, and Travel campaigns running Target CPA or Target ROAS. It does not apply to Display, App, Hotel, or video campaigns.
Microsoft, October 1: No Max CPC on New Automated Campaigns
Starting October 1, you cannot set a Max CPC — a ceiling on what one click can cost — when you create a new non-portfolio campaign using Maximize Conversions, Maximize Conversion Value, or Maximize Clicks. Campaigns created before that date that already have a cap keep it. Portfolio bid strategies keep it. So do Target Impression Share and enhanced CPC.
Microsoft's stated reason, in its own announcement, is that a Max CPC "provides conflicting instructions to the system." It wants you steering with budgets, targets, conversion value rules, and seasonality adjustments instead.
ChatGPT Ads: Automated Bidding Is Now the Default
OpenAI's "Maximize results" strategy is pre-selected for eligible new ad groups. It spends your budget chasing volume and makes no promise about CPA, CPC, or ROAS. If you want a ceiling, you switch to Manual: Max bid yourself. Nothing stops you. Nothing prompts you, either.
Why Google's Change Shows Up as Fewer Leads, Not a Bigger Bill
This is the part that trips people up, so it is worth being blunt about.
Your spend does not rise. Your daily budget still caps it. What falls is the number of leads that budget buys, because each one now costs closer to the number you told Google you would pay.
So if you watch spend to know whether something changed in your account, you will see nothing. Nothing moved. The account looks identical right up until you count leads at the end of the month and they are down — and by then you are a month into wondering whether it was seasonality, or the market, or a competitor.
Count leads and cost per lead. Those are the two numbers this change touches. Impressions and clicks will not tell you anything useful here, which is true most months and especially this one.
Then give it time before you judge it. Google suggests allowing one or two conversion cycles, which puts a defensible read at roughly 30 to 60 days after August 17. That lands mid-September through mid-October, which is about now for the earliest accounts. One caution: several platform changes landed in the same few weeks, so be careful about crediting this one for everything you see.
If You Set a Loose Target on Purpose, That Number Is Now a Bill
Plenty of accounts carry a target CPA set well above what the account actually delivers. Usually that was deliberate. A loose target gives automated bidding room to find conversions rather than choking on a number it cannot hit, and setting one was a reasonable thing to do.
It is now the opposite of reasonable. A number you never expected to pay is a number Google will spend up to.
So go and look. Pull every campaign on Target CPA or Target ROAS and compare the target you set against what the campaign actually delivered in the 90 days before August 17. If the gap is wide, that is a decision to make this week, not a surprise to explain next month. If the gap is narrow, you are fine and you just bought yourself some peace of mind for twenty minutes of work.
One honest caveat about fixing it. Changing a bid target by a meaningful margin puts the bid strategy back into learning — the settling period where delivery is unstable and costs are at their least predictable. So do not nudge it four times in three weeks. Move it once, to something near what the account really delivers, then leave it alone for a conversion cycle or two before judging the result.
Not sure whether your target is loose?
Send a screenshot of your campaigns view to saguarosocialmediaaz@gmail.com — the one showing target CPA next to actual cost per conversion is enough. I will tell you whether the gap is a problem and what I would set it to. One paragraph back, no pitch, and I will not follow up unless you ask me to.
What Microsoft's October 1 Deadline Actually Means
For most small advertisers, honestly, not much. Microsoft is usually a small slice of a small budget, and if you do not run it, this section is not your problem. Skip to what still controls your costs.
If you do run it, two things are worth saying.
First, the obvious move is a bad one. You could build a stack of Microsoft campaigns before October 1 purely to grandfather in a Max CPC. Do not. A campaign built to hold a setting is not a campaign built to perform, and you would be maintaining it for years for the sake of one checkbox.
Second, if Max CPC is the only guardrail holding your Microsoft account together, removing it exposes whatever was underneath. Every replacement Microsoft names — budgets, target CPA, target ROAS, conversion value rules, seasonality adjustments — assumes the platform can see your conversions. Most of them are useless without conversion tracking that works.
That is the actual project, and it is not a checkbox. Three weeks is enough time to get conversion tracking right on that account, and it is a far better use of them than pre-building campaigns.
ChatGPT Ads Defaults the Same Way
Short section, because ChatGPT Ads is early and honest advice about it is mostly about restraint.
There are no meaningful small-business benchmarks for it yet. That is not a knock on the platform, it is just how new an ad product it is. Run it as a test budget alongside something already proven, never as a replacement for it, and keep the test small enough that a null result does not hurt. The case for concentrating a small budget applies here more than anywhere.
And check the bid strategy on every new ad group you build. "Maximize results" being pre-selected is not a decision you made.
What Still Controls Your Costs
Three things, in the order they matter.
Your Budget
The only hard ceiling left. It is the one number on this page that means exactly what it says on every platform, and it did not change. If your worry is "what stops this from running away," the answer is still the budget, same as it was in July.
Your Target
The number the platforms now take literally rather than treating as a hopeful maximum. That makes it the most important setting in your account, and the one most likely to have been set carelessly — because until three weeks ago, being careless with it was mostly free.
Whether Your Conversion Data Is Telling the Truth
Every change on this page moves control from a manual cap to a target, and a target is only as good as the conversions feeding it. If your tracking counts a form view as a lead, double-counts phone calls, or misses half of them, the platform now spends confidently against a number that is wrong. It will hit the target you set. It just will not be measuring the thing you care about.
Tracking before spending was always the rule here. It just got more expensive to ignore. If you are not certain what your account counts as a conversion, that is the first thing to check, ahead of anything else on this page. What tracking should actually include covers what a working setup looks like, and it is the same list whether you do it yourself or pay someone.
What to Do Before October 1
Five things, in order, and the first one is the one people skip.
- Confirm what your account counts as a conversion. Submit a real test form or place a real test order and watch it register. Everything below depends on this being right.
- Audit every Target CPA and Target ROAS campaign. Compare the target you set against what the campaign delivered before August 17. Write the gaps down.
- Fix the loose targets once. Move each one to something near real delivery, then leave it alone for a conversion cycle or two.
- Get Microsoft conversion tracking working — only if Microsoft is in your mix. You have until October 1 to have something better than a Max CPC holding that account together.
- Check the bid strategy on any new ChatGPT Ads ad group — again, only if you run there. The default is volume, not a cost ceiling.
Items one through three apply to everyone running Google Ads, and none of this requires new spend. It is an afternoon of looking at settings you already have, which is the cheapest kind of work there is. What an afternoon like that costs if you would rather not do it yourself is on the pricing page, in full.
Common Questions About the Bidding Changes
Five worth answering directly.
Will My Ad Spend Go Up Because of Google's Change?
No. Your budget still caps what you spend. What changes is what you get for it: more expensive conversions, and fewer of them, for the same money. Watch lead count and cost per lead, not spend.
I Do Not Have a Bid Cap Set. Does This Affect Me?
The Google change does, and it is the one that matters most. It applies to any budget-limited Search, Shopping, Performance Max, Demand Gen, or Travel campaign running Target CPA or Target ROAS, whether or not you ever set a manual bid. The Microsoft and ChatGPT Ads changes only apply if you run on those platforms.
Should I Build Microsoft Campaigns Before October 1 to Keep My Max CPC?
No. Campaigns created before October 1 that already use Max CPC do keep the setting, but a campaign built to hold a setting is not a campaign built to perform. If Max CPC is your only guardrail, the real project is conversion tracking.
How Long Before I Know Whether Google's Change Hurt My Account?
Roughly 30 to 60 days, following Google's own guidance to allow one or two conversion cycles. Be careful about crediting this one change for everything, because several platform changes landed in the same few weeks.
Is There Any Bid Cap Left Anywhere?
Yes, in specific places. On Microsoft: portfolio bid strategies, Target Impression Share, and enhanced CPC. On ChatGPT Ads: switching to Manual: Max bid. On Google: manual CPC bidding still exists, but it means setting and adjusting a bid on every keyword by hand, which is more upkeep than most small accounts can sustain.