Why Your Performance Max Campaigns Are Burning Budget (And How to Fix It)

If you are running Performance Max (PMax) campaigns to scale an e-commerce or lead-gen brand across the GCC or Europe, you’ve likely noticed a frustrating trend: spend climbs rapidly, but your actual bottom-line revenue flatlines.

PMax is an incredibly powerful automation layer, but without strict guardrails, Google’s AI defaults to path-of-least-resistance spending. It chases cheap, low-intent impressions across the Google Display Network, junk mobile apps, and irrelevant YouTube placements just to hit its daily budget target. For cross-border businesses operating between tech hubs like Poland and high-competition markets like Dubai, this untamed automation is a massive drain on profitability.

The hidden bins where your money is disappearing

• The Display Network trap: By default, PMax bundles Search, Shopping, YouTube, and Display together. If your product feed or asset groups aren’t perfectly optimized, the algorithm dumps a huge percentage of your budget into cheap mobile app clicks that rarely convert.
• Brand cannibalization: Without an explicit brand exclusion list, PMax loves to bid on your own brand name. It looks like you have an incredible Return on Ad Spend (ROAS), but in reality the campaign is just stealing credit for high-intent organic traffic that would have converted anyway.
• Unoptimized asset groups: If you give Google text headlines but no high-quality video assets, Google will auto-generate low-quality slideshow videos for YouTube on your behalf. These look deeply unprofessional and fail to resonate with premium audiences in competitive UAE markets.

The 42% ROAS lift

The challenge: A lifestyle e-commerce brand operating in Poland wanted to expand into the UAE market. Their in-house PMax campaigns were generating sales but bleeding margin on broad, non-converting display placements across regional blogs.

The action: We implemented advanced script monitoring to isolate placement data, built rigorous negative placement lists, and separated asset groups by specific product margins rather than generic categories.

The result: Within 45 days, we eliminated 34% of wasted ad spend, shifting that budget directly into high-intent search and shopping placements — resulting in a 42% increase in blended ROAS across their Dubai and Warsaw operations.

The lesson: PMax is only as good as the guardrails around it. Feed hygiene, placement exclusions, brand exclusions, and margin-based asset groups are the difference between automation that compounds profit and automation that quietly spends it.