If you’ve been running Google Ads for a while, you know Performance Max has been the default champion—running the show, getting priority, and making advertisers adjust their strategies around it. But in late 2024, Google introduced major changes that shook up how PMax works, making control and strategy more important than ever.
So, we did what we do best—we ran a test. The goal? To push PMax beyond its limits and see what happens when you power it with smarter data. The results? Well, let’s just say PMax works a whole lot better when Google is fed with the right signals.
But before we get into the test results, let’s break down how Amanda AI fine-tunes PMax to maximize performance.
How it works: using Search data to fuel Performance Max
Performance Max is powerful, but let’s be real—it can feel like a black box. You feed it assets, audiences, and budget, and Google does its thing. But at Amanda AI, we make Google a whole lot smarter. How? By adding Search data into the mix.
Here’s how we fine-tune Performance Max for better results:
🚀 Integrating Search data – We pull audiences and keywords from Search and use them as signals in PMax asset groups. That means more precise targeting and less guesswork.
🤖 Automated asset group testing – Amanda AI dynamically configures and tests up to 99 asset groups, swapping out audience signals and search themes until we find the top-performing combinations.
🎯 Smarter audience signals – We refine audience targeting by diversifying signals, reducing overlap, and integrating new search themes to keep campaigns fresh and efficient.
⚡ Auto-Signals to Google – Our machine continuously updates and tests signals in audience groups, ensuring ongoing optimization without manual adjustments.
💰 Budget Reallocation – We don’t let money sit where it’s not working. Budgets shift automatically toward high-performing days, audiences, and segments—constantly optimizing to maximize results.
The test: powering Performance Max with a ton of Search data
We set up a test at the start of 2024 with a global company selling over 50,000 products across 8 markets. Before working with Amanda AI, they had been running 16 different Performance Max campaigns—but let’s just say the results weren’t exactly mind-blowing.
Here’s how their data looked before the test:
📌 16 different PMax campaigns
📌 Performance spread across multiple goals
📌 Budget distribution all over the place

The 16 different campaigns show performance in conversions.
Blue: Brand
Red: Performance Max
So, we stepped in and did our thing.
What did we do?
We fed Performance Max a ton of Search data, dynamically created highly targeted asset groups, and refined audience signals. Meanwhile, our machine constantly reallocated the budget to the highest-performing areas—running up to 5 million optimizations per day.

Before allowing us to run the test, the company tested Performance Max over an extended period of time for 16 different campaigns. The results were not mind-blowing.
The Results: a PMax strategy that actually works
✅ +23% increase in CTR – More precise audience targeting led to better engagement.
✅ €0.09 lower CPC – Smarter bid strategies meant lower costs.
✅ +89% increase in ROAS – Need we say more?

There’s more to the story, though. Budget distribution became more stable compared to their previous campaigns, and ad structure improved significantly, ensuring the ad spend worked harder than ever before.
The old setup was leaking efficiency—too many campaigns, too many scattered budgets, and not enough control over how Performance Max was running. By refining Performance Max with search data, asset testing, and audience optimization, Amanda AI cleaned up the inefficiencies and delivered better results.

Blue: Amanda AI’s Performance Max campaign shows the number of conversions (running only one campaign.)

The budget distribution with Amanda AI was more stable than that of previous campaigns.
Blue: Amanda AI budget distribution
What this means for you as an advertiser
If you run Performance Max without a clear strategy, there’s a big chance you’re wasting your marketing budget. It’s not about setting up campaigns and leaving it at that—it’s about feeding Performance Max the right data, refining audience signals, and making sure budgets are working at full efficiency.
At Amanda AI, that’s exactly what we do. We take Performance Max beyond its default settings and turn it into a high-performance machine.
The takeaway? An automated service like Amanda AI can really help an automated campaign type like Performance Max work harder and smarter. Just because it has automation at its core doesn’t mean it should be left alone to do its own thing; rather, it should be fed more data points to outperform itself.
In other words, we help Performance Max outperform Performance Max.
Curious to see what smarter campaign management looks like? Let’s talk.