Quick Answer: The strongest version of this claim is that Google's PMax reporting is structurally designed to make the campaign look better than it is -- by hiding placement data, counting view-through conversions by default, siphoning branded traffic, and burying attribution problems inside a "black box" that advertisers can't audit. Whether that's gaslighting or just aggressive product design depends on your perspective. What isn't debatable is that the default reporting paints a rosier picture than the underlying business reality.
What "Gaslighting" Means in This Context
Gaslighting, in the advertising context, means this: Google shows you numbers that look like strong performance, and when you raise concerns, you're shown more numbers that look like strong performance. The metrics appear to validate the product. But when you look at your CRM, your actual customer count, your pipeline -- the numbers don't match.
You're not crazy. You're just looking at a dashboard that was designed to show you a specific story.
The Four Ways PMax Reporting Flatters Itself
1. It hides where your ads actually ran
PMax doesn't show you a clean placement report the way Display or Standard Shopping does. You can see aggregated insights, but the full breakdown of which specific sites, apps, and placements received your budget -- and at what cost -- isn't available in the standard interface. The only way to see it is through scripts, third-party tools, or a channel insights report that shows high-level distributions, not individual placements.
This isn't a neutral design choice. If you could see that 40% of your budget went to mobile app placements and made-for-adsense sites with high bot conversion rates, you would act on that information. The opacity protects Google's ability to monetize that inventory.
Google hides the data intentionally so they can make more money. The scripts are all best guesses -- even the scripts can't access the accurate data because Google hides it.
2. It counts view-through conversions by default
Performance Max reports view-through conversions alongside click-through conversions in its totals. View-through conversions are recorded when someone sees your ad but doesn't click, and then later completes a conversion through any other channel.
The conversion window for view-through can be up to 30 days. During that window, if someone converts through organic search, a direct visit, or a branded search -- PMax gets credit. The impression becomes a conversion. The campaign looks efficient.
A Google rep acknowledged this privately: "most PMax conversions are view-through when not brand included."
This information is technically disclosed in the Google Ads interface. But it's not the number Google leads with in campaign reports. The default presentation is total conversions. View-through is a column you have to add manually.
3. It competes with your branded campaigns and claims credit for the results
When PMax runs without brand exclusions, it bids on queries from people who are already searching for you. It wins impressions from your branded Search campaign. It converts those visitors and reports them as PMax conversions.
The business reality is unchanged -- those customers already knew who you were. But PMax's conversion total went up, your account-level ROAS held steady, and from the dashboard's perspective everything looks fine.
PMax hijacking brand and pretending it's incremental is infuriating. It's basically paying yourself for clicks.
4. It optimizes toward whatever you measure, not what you actually want
Performance Max cannot distinguish between a bot filling out a form and a real qualified buyer doing the same thing. If your conversion action is a form submission, PMax optimizes toward form submissions. It finds the cheapest possible way to produce them, which includes low-quality placements and bot traffic.
pMax is one big data whore that only cares for a few things: spending max budget on all channels, getting more data for smart bidding, clicks.
When you report poor lead quality, the response from Google -- through reps, through optimization suggestions, through the interface -- is often to give PMax more budget, more conversion data, more time. More of the same inputs, promised to produce better outputs. That's a hard loop to question from inside a product that controls its own reporting.
What Makes This Hard to Prove
None of the four mechanisms above are hidden from the terms of service or technically false in the way a lie would be. Google discloses that view-through conversions exist. The Auction Insights report will show you if branded impression share is dropping. The placement insights tab shows some channel data. The information is technically available.
The problem is the default. By default, you see a clean conversion total, a strong ROAS, and a recommendation to increase budget. You have to know what questions to ask and where to look to find the counterevidence. Most advertisers -- and many agencies -- never ask.
Google's reps know the truth and stay quiet. You have to know to ask -- and even then, you get it quietly, not in writing.
What Honest PMax Measurement Looks Like
If you want to know whether your PMax campaign is actually doing what it claims, you need to build your own evidence outside Google's reporting:
Reconcile against your CRM. Compare Google Ads-reported conversions to confirmed customers or qualified leads in your CRM. A 50% gap is a problem worth investigating. A 90% gap means the campaign isn't doing what the dashboard says.
Segment click-through vs. view-through. Add the view-through conversion column to your reports. If view-through represents more than 20-30% of your totals, the reported ROAS is significantly inflated.
Run a brand exclusion test. Exclude brand from PMax for four weeks and compare total revenue vs. the prior period. If PMax conversion volume drops significantly but revenue holds, you've confirmed cannibalization.
Do an incrementality test. Run PMax in one geographic market and pause it in a comparable market. Compare total revenue from both. The gap between them is your best estimate of what PMax is actually generating vs. what would have happened anyway.
Is Any of This Malicious?
Probably not in the conspiratorial sense. Google is a publicly traded company optimizing for its own revenue, which comes from ad spend. PMax's defaults (view-through attribution, brand bidding, placement opacity) all increase spend and make the product look better. That's predictable corporate behavior, not a secret campaign against advertisers.
But the practical effect is the same: the product is structured to be trusted at face value, and the tools required to verify that trust are mostly unavailable, obscured, or require significant technical investment to build. Advertisers who run PMax without auditing their own data are almost certainly overestimating its value.
FAQs
Should I stop using Performance Max entirely?
Not necessarily. PMax can generate genuine incremental performance for accounts that exclude brand, use offline conversions as the optimization target, and run proper placement monitoring. The issue isn't the campaign type -- it's running it with default settings and trusting default reporting.
My Google rep told me PMax is performing great and I should increase budget. Should I trust that advice?
Google reps are measured on spend. That doesn't make them dishonest, but their incentives don't align with yours. Before taking advice to increase budget, reconcile your CRM against reported conversions and run a view-through vs. click-through breakdown. Let your own data answer the question.
Are there third-party tools that can see PMax placement data?
Some scripts and third-party platforms (Optmyzr, Mike Rhodes' PMax scripts) can surface more placement detail than the native interface exposes. They're useful but imperfect -- Google's data access limits mean even scripts are working with aggregated or sampled data, not the full placement log.
If I exclude brand and view-through conversions, my ROAS will look terrible. How do I explain this to clients?
The explanation is that the previous ROAS was measured incorrectly, and this is what the campaign actually produces. That's a harder conversation than showing declining numbers, but it's an honest one. Clients who understand attribution will appreciate it. Clients who don't may need a broader education conversation first.