Performance Max and Click Fraud: The Hidden Problem Google Won't Talk About

Google loves Performance Max. They’ve spent the last three years pushing every advertiser towards it — from enterprise brands to one-person plumbing businesses. The pitch is compelling: one campaign type, across every Google property, powered by AI that optimises everything for you.

What Google doesn’t talk about is the fraud problem.

Performance Max campaigns operate as a black box. You can’t see exactly where your ads are shown. You can’t see detailed information about who’s clicking them. And you have far less control over where your budget goes compared to traditional Search campaigns. For fraudsters — both human and automated — this lack of transparency is a gift.

This article explains how Performance Max exposes your budget to click fraud, why Google’s own protections fall short for PMax specifically, and what you can do to protect your ad spend.

What Is Performance Max (and Why Does Google Push It So Hard)?

If you’re not familiar with it, Performance Max (often shortened to PMax) is a campaign type that runs your ads across all of Google’s advertising networks simultaneously: Search, Display, YouTube, Gmail, Discover, and Maps. Instead of you choosing where your ads appear and what they look like, Google’s AI handles most of those decisions for you.

Google positions this as a major benefit — less work for you, more reach for your budget. And to be fair, PMax can work well for some businesses. The AI is genuinely good at finding conversions when it has enough data to learn from.

But there’s a reason Google pushes PMax so aggressively, and it’s not purely altruistic. Traditional Search campaigns only run on Google Search — one channel. PMax runs across Google’s entire advertising ecosystem, including the Display Network and YouTube, where ad inventory is vastly more abundant (and cheaper for Google to fill). More inventory served means more revenue for Google, regardless of whether those impressions and clicks are valuable to you.

This matters for fraud because Display Network and partner site placements have historically had dramatically higher fraud rates than Google Search. When your budget is spread across all these networks automatically, a significant portion can end up on placements where fraud is rampant — and you might never know it.

The Transparency Problem: A Black Box for Your Budget

The fundamental issue with Performance Max and fraud comes down to visibility — or rather, the lack of it.

What you can’t see

With a traditional Search campaign, you know exactly which keywords triggered your ads, which search terms people typed, and you can review every element of performance in detail. With PMax, Google provides far less granular data.

You get aggregated performance metrics — total clicks, conversions, cost — but the details of where those clicks came from are limited. Google has improved PMax reporting since its launch, adding placement reports and some asset-level data, but the level of detail still falls well short of what you’d get from a standard Search campaign.

This means that if a chunk of your PMax budget is being consumed by fraudulent clicks on Display Network placements or dubious partner sites, the aggregated reporting makes it very difficult to spot. The fraud is hidden inside a bundle of metrics that look acceptable on the surface.

The “Made for Advertising” site problem

One of the most well-documented fraud vectors within PMax relates to MFA (Made for Advertising) sites. These are websites created specifically to host ads and generate revenue from clicks, with little or no genuine content or audience. They exist purely to siphon advertising budgets.

Research from the Association of National Advertisers found that a significant share of programmatic ad impressions end up on these junk sites. When PMax’s algorithm places your ads on Display Network sites, some of those placements inevitably land on MFA sites where the “visitors” are bots, click farms, or incentivised clickers generating fraudulent engagement.

You’re paying for clicks that come from sites nobody real is visiting, on pages that exist only to steal your budget. And because PMax doesn’t give you granular placement control, you can’t easily opt out of these placements the way you could with a manually managed Display campaign.

How Click Fraud Works Differently on PMax

Click fraud on Performance Max campaigns has some unique characteristics that make it particularly damaging.

The algorithm poisoning problem

This is the most insidious issue, and it’s one that most advertisers don’t understand.

PMax’s AI learns from your campaign data. When someone clicks your ad and then converts (calls you, fills in a form, makes a purchase), the algorithm learns what that type of visitor looks like — their location, device, browsing behaviour, time of day, and dozens of other signals. It then optimises to find more people like them.

But here’s the problem: when bots or fraudulent clickers interact with your ads, they pollute the data the algorithm is learning from. If bot traffic comes from certain networks, device types, or geographic patterns, the AI starts “learning” that these patterns are valuable — and actively seeks out more of the same.

This creates a vicious cycle. Fraudulent clicks feed the algorithm bad data. The algorithm optimises towards that bad data. More budget goes to fraudulent placements. More bad data comes in. Your campaign performance degrades steadily, and the cause is almost invisible because it’s happening inside Google’s black-box optimisation.

With a traditional Search campaign, you can see exactly which keywords are performing and cut the ones that aren’t. With PMax, the optimisation happens behind a curtain, and by the time you notice performance declining, the algorithm may have already been significantly poisoned.

The cross-network amplification effect

Because PMax runs across every Google network simultaneously, fraud on one network can drag down performance across all of them. If fraudulent Display clicks poison the algorithm’s learning data, it doesn’t just affect Display placements — it can shift budget allocation across all channels, reducing your effective presence on Search (where the highest-intent, most valuable clicks typically come from).

In a standard campaign structure, you’d have separate budgets for Search and Display, and fraud on Display wouldn’t directly cannibalise your Search budget. PMax removes that separation, meaning a fraud problem on one network becomes a budget problem everywhere.

Competitor clicking is harder to detect

Competitor click fraud is common in local markets. With a Search campaign, you can at least analyse time-of-day click patterns, geographic data, and session behaviour to spot suspicious activity. With PMax, the reduced reporting granularity makes these patterns much harder to identify.

A competitor clicking your PMax ads might trigger clicks on your Search listings, your Display placements, or your YouTube ads — and because PMax bundles all of this data together, the suspicious pattern gets diluted across multiple channels. What would be an obvious spike in a Search-only campaign becomes statistical noise in PMax’s aggregated reporting.

The Numbers: How Much Is PMax Fraud Costing You?

Quantifying PMax fraud precisely is difficult because, by design, the data isn’t transparent enough to measure easily. But the evidence from industry research paints a concerning picture.

Studies have found that Smart campaigns (the predecessor to PMax that shared many of the same automation features) had significantly higher invalid traffic rates compared to standard manually managed campaigns. The combination of automated placement selection, reduced advertiser control, and broad network coverage creates ideal conditions for fraud.

For Display Network placements specifically — which are a major component of PMax — fraud rates have consistently been documented as much higher than pure Search traffic. When your PMax campaign allocates a portion of your budget to Display (as it inevitably will), that portion faces elevated fraud risk.

If you’re spending £3,000 per month on PMax and even 15–20% of your clicks are fraudulent, that’s £450–£600 per month in wasted budget. Over a year, that’s £5,400–£7,200 that generated zero leads and actively damaged your campaign’s AI optimisation. And in competitive industries like home services, legal, and dental — where CPCs are high and competitors have strong incentives to click your ads — the percentage can be even higher.

To estimate the fraud exposure for your specific PMax campaigns, try our ROI Calculator — enter your monthly PMax spend, average CPC, and industry to see what the numbers look like.

Why Google’s Invalid Click Protection Falls Short on PMax

Google does have automated invalid click filtering that runs across all campaign types, including PMax. This system catches some obvious fraud — known bot signatures, data centre IP addresses, rapid repeat clicks from the same source.

But Google’s filtering has specific weaknesses that are amplified by PMax’s structure.

The incentive problem. Google earns revenue from every click, including clicks on Display Network placements and partner sites. Their invalid click system needs to balance protecting advertisers with maintaining the revenue those clicks generate. For a detailed analysis of this dynamic, see our click fraud statistics page.

Cross-network complexity. PMax’s multi-network structure means fraud signals that might be clear on a single channel get fragmented across many channels. A pattern of suspicious clicks that would be obvious in a Search-only campaign might not trigger filters when diluted across Search, Display, YouTube, and Gmail simultaneously.

Real-device fraud. Google’s filters are strongest against bot fraud from data centres and known bad actors. But competitor clicking from real devices on residential networks — one of the most common fraud types in local markets — looks like legitimate traffic to Google’s automated systems. On PMax, where you can’t easily investigate individual clicks, this type of fraud has even more room to operate undetected.

Limited retroactive credits. Even when Google does identify invalid clicks after the fact and issues credits, these credits rarely reflect the full extent of the waste. And they don’t address the algorithm poisoning problem — by the time a credit arrives, the damage to your campaign’s optimisation data has already been done.

What You Can Do to Protect Your PMax Campaigns

Despite the challenges, there are concrete steps you can take to reduce fraud exposure in your PMax campaigns.

Step 1: Review your placement reports

Google now provides placement reports for PMax campaigns showing where your ads appeared on the Display Network and YouTube. Review these regularly and look for placements on sites you don’t recognise, sites with suspicious names (strings of random characters, generic domains), or sites that seem entirely unrelated to your business.

You can create placement exclusion lists in Google Ads and apply them to your PMax campaigns. While this won’t catch everything, it removes the most obvious junk placements from your rotation.

Step 2: Monitor channel-level performance

Within PMax reporting, you can see a broad breakdown of performance by channel (Search, Display, Video, etc.). If Display is consuming a disproportionate share of your budget relative to the conversions it generates, that’s a signal worth investigating. Some advertisers find that opting out of Display-heavy placement groups through asset group restructuring can reduce fraud exposure.

Step 3: Use audience signals aggressively

PMax allows you to provide “audience signals” — essentially hints to the algorithm about who your ideal customers are. The stronger and more specific your audience signals, the less likely the algorithm is to wander into low-quality traffic pools. Use your customer match lists, website visitor data, and detailed demographic targeting to guide the AI.

Step 4: Run PMax alongside standard Search campaigns

Many experienced PPC managers don’t rely on PMax alone. They run traditional Search campaigns for their highest-value keywords (with full control and visibility) and use PMax as a supplementary campaign for incremental reach. This way, your core lead-generating keywords have the transparency and control of standard campaigns, and PMax handles the broader discovery without being your only campaign type.

This approach also gives you a control group — if your Search campaign metrics are steady but your PMax metrics are degrading, you can isolate the PMax issue more easily.

Step 5: Connect Google Analytics 4 for deeper analysis

Link your GA4 account to your Google Ads account if you haven’t already. While PMax’s native reporting is limited, GA4 can show you engagement metrics for the traffic PMax sends to your website — session duration, pages per visit, bounce rate, and conversion events. If PMax traffic shows consistently lower engagement than your Search traffic, that’s worth investigating.

Step 6: Use automated click fraud protection

Manual monitoring has real limits with PMax. The black-box nature of the campaign type means that many fraud signals are only visible at the click level — device fingerprints, behavioural patterns, network characteristics — rather than in aggregated reporting.

Automated click fraud protection tools like ClickGuardian monitor every visitor that arrives on your site from PMax (and all other campaign types), analysing behavioural signals in real time. When a visitor shows patterns consistent with fraud — zero engagement, device fingerprint matches with known fraudulent sources, repeat visit patterns across different IPs — the source is blocked before it can waste more budget or further pollute your algorithm’s data.

This is particularly valuable for PMax because it addresses the algorithm poisoning problem at its source. By blocking fraudulent visitors before they register as “data” in your campaign’s learning model, you keep the AI optimising on clean signals rather than garbage data.

Should You Stop Using Performance Max?

Not necessarily. PMax can be effective, particularly for businesses with sufficient conversion data for the algorithm to learn from (Google recommends at least 15–30 conversions per month). The issue isn’t that PMax is inherently bad — it’s that its lack of transparency creates fraud vulnerabilities that advertisers need to actively manage.

The worst approach is to set up PMax, leave it running, and assume Google’s built-in protections are handling everything. That’s how budgets get quietly drained over months without anyone noticing.

The smart approach is to use PMax with your eyes open: monitor placement reports, track channel-level performance, run it alongside standard Search campaigns for comparison, and add a layer of automated click fraud protection to catch what Google’s filters miss.

For home services businesses, legal practices, dental clinics, and other local advertisers in competitive markets, PMax fraud protection isn’t optional — it’s the difference between a campaign that generates leads and one that feeds an algorithm on junk data.

Don’t let PMax drain your budget — see how much you could save with proper protection.

Frequently Asked Questions

Is Performance Max more vulnerable to click fraud than Search campaigns?

Yes, for several reasons. PMax runs across all Google networks including Display, which has higher fraud rates than Search. It provides less granular reporting, making fraud harder to detect. And its automated optimisation can be poisoned by fraudulent click data, creating a feedback loop that worsens over time. Standard Search campaigns give you more visibility and control to identify and address fraud.

Can I see where my Performance Max ads are being shown?

Google has improved PMax placement reporting, and you can now see where your ads appeared on Display Network and YouTube. However, the reporting is still less detailed than what standard campaigns provide, and you can’t see the same level of search term data for your Search placements within PMax. Check your placement reports regularly and build exclusion lists for suspicious sites.

Will Google refund me for fraudulent clicks on PMax campaigns?

Google’s invalid click filtering applies to PMax just as it does to other campaign types, and you won’t be charged for clicks Google identifies as invalid. However, Google’s filters are less effective at catching sophisticated fraud — particularly from real devices and residential networks — and the refund process doesn’t address the damage fraudulent clicks cause to your PMax algorithm’s learning data. For more on Google’s limitations, see our click fraud statistics.

Should I switch back to standard Search campaigns?

For your highest-value keywords, running standard Search campaigns alongside PMax is a strong strategy. This gives you full control and transparency for the terms that generate your most important leads, while PMax can handle incremental reach. Many PPC professionals recommend this hybrid approach rather than relying entirely on PMax. See our complete guide to stopping click fraud for campaign-level protection strategies.

How does ClickGuardian protect PMax campaigns specifically?

ClickGuardian monitors every visitor that arrives on your site from any campaign type, including PMax. It analyses behavioural signals, device fingerprints, and network patterns in real time to identify fraudulent visitors. By blocking bad traffic before it registers as conversion data in your PMax campaign, it prevents the algorithm poisoning problem — keeping your AI optimisation learning from real customer behaviour rather than bot activity. Industry-specific protection information is available for plumbers, HVAC, roofers, lawyers, and all industries.

---

Last updated: March 2026. For the latest click fraud data and industry benchmarks, see our Click Fraud Statistics page.