The 'Performance Paradox': Why Your PMax Campaigns Need a Surgical Intervention

Right, let's cut to the chase. Performance Max (PMax). Google's answer to the all-encompassing, automated advertising machine. For many of you – marketing managers, founders, and the unsung heroes in in-house teams – it probably sounded like the dream: one campaign, all channels, maximum conversions. What you're likely experiencing, however, is a black box that’s increasingly difficult to decode, often chewing through budget with questionable returns, especially for anything beyond straight-up e-commerce.

It’s 2026, and the honeymoon period with PMax is well and truly over. We're beyond the 'just feed it good assets' advice. What we’re seeing, and what you should be seeing, is a necessity for surgical, precise interventions. PMax isn't a magic wand; it's a powerful but sometimes reckless tool that needs constant calibration to perform optimally without cannibalising other efforts or wasting spend. If you're not getting granular, you're just throwing money at Google's algorithm and hoping for the best. And 'hope' isn't a strategy.

The Unseen Cannibalisation: Where's Your Budget Really Going?

This is perhaps the biggest headache with PMax, especially if you run strong, branded search campaigns. PMax loves to swoop in and take credit for branded conversions that would have likely happened anyway, often at a lower CPA through your established search campaigns. It’s like a hungry teenager raiding the fridge – easy wins, minimal effort, and your carefully stocked snacks (read: budget) disappear.

Without proper exclusion strategies and meticulous monitoring, PMax can inflate perceived performance while deflating actual incrementality. We've seen clients with PMax campaigns showing fantastic 'results' on paper, only to find a significant portion were simply capturing existing demand. You need to ask yourself: would this conversion have happened regardless of PMax?

Asset Group Audits: Beyond 'Good Enough'

The quality and variety of your asset groups are paramount. We're not just talking about having 'enough' images and headlines. We're talking about a continuous, analytical approach to optimisation. Treat each asset group as its own micro-campaign.

• **Specificity is Key:** Don't throw all your products or services into one generic asset group. Segment aggressively. If you sell luxury watches and budget smartwatches, they need separate asset groups with tailored messaging and audience signals.
• **Creative Freshness:** Google's algorithms learn fast, and asset fatigue is real. Rotate your creatives regularly – weekly, even bi-weekly for high-spend campaigns. Think A/B/C testing for images, videos, and headlines.
• **Performance-Driven Refinement:** Utilise the 'Combinations' report within PMax. It’s not perfect, but it gives clues. More importantly, pull asset-level reporting (via scripts or third-party tools if necessary) to identify underperforming assets and replace them ruthlessly. If a video has a terrible view-through rate, get rid of it.

Leveraging Audience Signals (Not Just 'Feeding' Them)

Google tells you to 'feed' PMax with audience signals. That's a good start, but it's not enough. You need to treat these signals as hypotheses to be tested, not just data dumps.

1. **Granular Custom Segments:** Go beyond generic in-market audiences. Build highly specific custom segments based on competitor URLs, specific high-intent keywords, and even YouTube channels your target audience consumes. Test different combinations.
2. **First-Party Data Reinforcement:** Upload every piece of first-party data you have – customer lists, past purchasers, abandoned cart users, newsletter subscribers. Refresh these regularly. The more Google understands your ideal customer, the better it can find similar audiences.
3. **Exclusion by Design:** Crucially, use your audience signals for *exclusions* too. If you’re targeting new customers, exclude existing ones. If you’re pushing a high-value service, exclude audiences known for low-value purchases.

The (Necessary) Art of Exclusion Management

This is where PMax really starts to bite if you’re not proactive. Google wants PMax to run everywhere, and that includes questionable placements or brand-damaging environments. You need to fight for your brand safety and budget efficiency.

• **Brand Safety First:** Request a brand exclusion list from Google support. This isn't optional; it's a fundamental step. Don’t wait for horror stories of your ads appearing next to unsavoury content.
• **Negative Keyword Sculpting:** While PMax largely limits negative keywords, work with your Google rep to get them applied at the campaign level. Prioritise generic, low-intent terms or brand terms you want to protect in other campaigns.
• **Placement Exclusions:** Monitor placement reports in your Google Ads account (often found under 'Where ads showed'). Identify low-performing or irrelevant placements and exclude them. This requires ongoing vigilance; PMax will always try to find new inventory.

Your PMax Isn't 'Set and Forget' – It's 'Set and *Surgically Intervene*'

Here's our blunt take: if you're treating PMax as a 'set and forget' campaign, you're effectively outsourcing your budget management to Google's algorithms, without the necessary human oversight. PMax *can* deliver incredible results, but only when paired with diligent, data-driven management. It demands continuous testing, ruthless optimisation of assets and audience signals, and a proactive stance on exclusions.

Stop hoping your PMax campaigns are performing; start *knowing* they are. This means diving deep, getting granular, and performing those surgical interventions that transform a budget-hungry black box into a precision-targeted conversion engine. Your ROAS – and your sanity – will thank you.