Google Performance Max and Meta Advantage+ now decide where your budget goes and report results in aggregate you cannot act on. They also take credit for conversions already in the funnel. Handing spend allocation to a self-reporting black box is the opposite of knowing your numbers, and here is how to keep enough control to still answer the only two questions that matter.
You cannot optimise what you cannot see. You cannot see where your money went. So what exactly are you optimising?
The Take: Google Performance Max and Meta Advantage+ now decide where your ad budget goes and hand back a report on how well they did. That is not automation, it is a black box marking its own test. Across New Rebellion's benchmark data, 21.7% of Australian businesses score below 50 on Data & Tracking, meaning most operators handing spend to these platforms have no independent instrument to check the platform's claim against.
I have worked both sides of this table. The client side, where you sign off the spend and then wait for someone to tell you it worked. The agency side, where you are the one writing the report that says it worked. So take it from someone who has watched this play out, this is an old trick wearing a new coat of paint.
The AI systems running your campaigns have taken the wheel. Performance Max and Advantage+ now choose where your dollars land, then hand back a summary telling you how it went. The system spends and the same system judges the outcome. A worrying share of advertisers never question it.
What exactly did you hand over when you turned on Performance Max or Advantage+?
Control over targeting, creative and budget, in exchange for one aggregate number covering all of it, a single figure you cannot break down by placement, audience or the 6+ channels these platforms span.
You handed over targeting, creative rotation and budget allocation. In return you got back a single aggregate number. You can't tell which placement worked, which audience actually bought or where the budget vanished. That fog is the design, not an accident.
Worth being precise about what these tools actually do, since the pitch decks sell a cleaner story than reality delivers. Meta's Advantage+ Shopping takes over targeting, creative rotation and budget pacing without asking. Launch the campaign and the system chooses the audience, the creative and the price paid to reach them. Google's Performance Max runs the same play across Search, Display, YouTube, Gmail and Maps under one campaign umbrella. Supply the assets and the objective. Get one summary figure in return.
The number is the problem. The results land aggregated. You get the whole-campaign total, never the component parts. There's no visibility into which slot delivered, which segment bought or where the wasted spend hid. Google cracked the lid open slightly in April 2026, adding asset-level disapprovals and limited demographic detail, though the top-line figure still lands as an undividable lump.
Can Australian businesses actually check what these platforms are telling them?
For most, no. New Rebellion has scored 723 Australian businesses across 70 industries on Data & Tracking, the dimension covering measurement infrastructure. 21.7% score below 50, notably weaker than the 10.9% who fall below 50 on Acquisition Performance.
Methodology: scores are drawn from New Rebellion's benchmark dataset of Australian businesses, scored 0 to 100 across six marketing dimensions including Data & Tracking. As at July 2026, n=723 AU businesses. The gap between the two rows above is the point of this piece. Businesses are weaker on the infrastructure needed to check a platform's claim than they are on the channel performance itself. See How We Score for the full methodology.
The gap matters. Roughly one in five Australian businesses stepping into Performance Max or Advantage+ lack an independent measurement layer capable of catching an inflated claim. They rely entirely on the platform's word, and the platform has every reason to report a flattering figure.
It's the same trap I watched play out for years on the agency side. Whoever runs the spend is incentivised to say it's working. Now the spend runs on autopilot and the incentive hasn't changed. The human left. The conflict of interest stayed.
Are these platforms actually finding new customers, or just claiming credit for sales that were already coming?
More often the latter than advertisers assume. Advantage+ Shopping will log a sale from a shopper already deep in your funnel and claim it as a conversion the system generated, and the default settings push a real chunk of the budget into retargeting people who were already reaching for their card before the ad appeared.
Consider what that costs commercially. The system chases your warm audience, claims a sale that was coming regardless and files itself a glowing result. The figure looks good, and you feel good reading it. The number itself is genuine. The credit attached to it is not.
The share of Australian businesses in New Rebellion's benchmark data who clear a score of 50 on Data & Tracking, the measurement layer that would catch a platform overstating its results. The other 21.7% have no real way to check the claim.
I have measured this market. Across Australian businesses the pattern is consistent and it is ugly. The ones who hand spend to the platform and read the platform's report cannot tell me what their money actually did. They can tell me the platform's number. Ask them how much of that conversion volume would have landed anyway, with no ad at all. The room goes quiet. That silence is the whole article.
What should you actually do to keep a real check on the spend?
Hold back a slice of the account you can read yourself. Exclude your existing customers and warmest retargeting pools where the platform allows it, and watch what the reported conversion number does.
This isn't an argument against automation. Worth being plain about that so the point doesn't get lost. The system handles certain tasks faster than any human team. It runs more test combinations in an hour than a person manages in a month. The tool isn't the problem. Giving up visibility and dressing it up as efficiency is.
Here's the practical version. Keep some structure in the account. Don't route every dollar through one automated campaign and call that a plan. Set aside enough spend into readable channels and campaigns to give yourself a control group that tells the truth. Turn the automated spend down. If revenue holds, that tells you something. If revenue drops, that tells you something else. Either way, the test costs you almost nothing to run.
Where the platform allows it, pull existing customers and your warmest retargeting pools out of the automated campaigns, then watch the conversion figure move. A collapse means the platform was booking sales you already had. If it holds steady, the platform is genuinely finding new buyers. There's no way to know without testing it, and the platform is never going to volunteer that answer.
Keep two questions front of mind throughout. How much money does the business make. How much does it spend. Everything else can be worked back from there, but without answers to those two from your own figures rather than the platform's, there's no strategy, only a hope.
Where does this leave Australian advertisers?
The money is moving toward these platforms, not away from them. Australian ad spend is forecast to grow 7.4% in 2026, led by retail media, search and social, more automation and more aggregate reporting arriving with it. That trend is not stopping and I am not asking you to stand in front of it.
I am asking you to keep enough of the wheel in your hands to know where the car is going. Let the platform drive the parts it drives well. But keep your own instruments running, keep a corner of the account you can read and never take the report at face value without checking it. An operator who does not know what their money did is not running marketing. They are funding someone else's report and hoping it is true.
If you want to see how your own measurement infrastructure actually scores, get your scorecard through Hub and find out before the next platform report lands in your inbox.
Frequently asked questions
Why can't I see which part of my Performance Max campaign is actually working?
Google Performance Max reports results in aggregate across Search, Display, YouTube, Gmail and Maps as a single number. Google opened limited asset-level and demographic reporting in April 2026, but the core conversion figure still can't be broken down by placement or audience.
Does Meta Advantage+ take credit for sales that would have happened anyway?
Often, yes. Advantage+ Shopping's default settings send a meaningful share of budget toward retargeting warm audiences who were already close to purchasing, and the platform reports those sales as conversions it drove.
How do I check if my ad platform's reported results are accurate?
Keep a portion of your account outside the automated campaigns so you have a readable control group. Exclude existing customers and warm retargeting pools from automated spend where the platform allows it, then watch whether the reported conversion number holds or collapses.
Should I stop using Performance Max and Advantage+ altogether?
No. These platforms are genuinely effective at testing combinations faster than a human team can. The problem isn't the automation, it's running 100% of spend through a black box with no independent way to verify what it reports.