...

Mastering Performance Max Campaigns: 2025 Updates & Best Practices

Picture of Author

Author

Gabriela

Read time

6 mins

Posted on

2 Jun 2025

story

How to Optimise Google’s Most Powerful AI Campaign Type for Maximum ROI

Performance Max campaigns have levelled up in 2025. From negative keywords to new customer bidding and asset group reporting, this guide walks you through every update – plus proven strategies to drive conversions with Google’s most automated campaign type.

Google’s Performance Max (PMax) campaigns have matured significantly, and 2025 brings new features that give advertisers more control and insight than ever before. Performance Max is an AI-driven campaign type that runs across all Google channels (Search, Display, YouTube, Gmail, Discover, etc.) from a single campaign, optimising in real-time to meet your goals. To truly master Performance Max in 2025, marketers need to understand the latest updates and apply best practices that align with Google’s automation and their own business objectives.

What’s New in Performance Max (2024–2025):
In response to advertiser feedback, Google rolled out important updates to Performance Max heading into 2025. Notably, more campaign controls and transparency tools have been introduced. For example, campaign-level negative keywords are now available to all advertisers. This means you can exclude specific search queries (like irrelevant or low-value terms) from your PMax campaigns – a welcome feature for brand protection and budget optimisation.

Google also launched a “new customer acquisition” goal with High Value mode. This allows businesses to bid more for new users likely to become high-LTV customers (using your data about customer lifetime value) – perfect for prioritising quality over quantity in customer acquisition.

Retailers running PMax with Merchant Center product feeds gained brand exclusions by format, so you can, for instance, stop PMax from showing text ads on your brand terms while still allowing Shopping ads for those terms. This granular control helps prevent cannibalising your own branded search campaigns.

Google has even been beta-testing demographic and device targeting controls (e.g. excluding 18–24 age or focusing on mobile only), signalling that more levers are coming to steer the AI.

On the reporting side, Search term insights are improved – Google’s “search themes” feature lets you indicate important keywords, and now you can see which queries came from your provided themes versus Google’s automation. Also, asset group reporting is more detailed, allowing performance breakdowns by asset group so you can identify top-performing creative combinations.

These updates collectively address many early criticisms of PMax by giving advertisers greater clarity and command.

Best Practices for PMax Success:

1. Feed the Right Data & Conversions
Accurate conversion tracking is fundamental for PMax. Since bids and placements optimise towards your conversion goals, ensure your Google Ads conversion actions are set up correctly and reflect your true business objectives. If you’re an e-commerce retailer, use conversion value (revenue) and consider importing offline sales if applicable. Use enhanced conversions or offline conversion imports to feed real sales data back into Google.

For lead-gen businesses, choose the most meaningful conversion (e.g. qualified lead or sale, not just raw form submissions) or use conversion value rules to weight leads by quality. PMax now allows setting new customer conversion goals too – use that if growing your customer base is a priority. In short, give the AI the best targets; it can only optimise for what you define as success.

2. Leverage New Controls Thoughtfully
With negatives and exclusions now available, use them strategically to avoid waste. Add your brand keywords as negatives in generic PMax campaigns to ensure budget is spent on new customers, not people searching your brand (which you might capture via cheaper brand search campaigns).

If you operate in multiple product categories, use “URL contains” rules or separate asset groups to direct PMax to relevant sections of your site (e.g. one PMax campaign for “/shoes/” pages, another for “/accessories/”). This can act like a rudimentary way to cluster products or services so the AI doesn’t mix completely unrelated offerings.

Consider using the new customer acquisition setting if your goal is finding fresh customers – in High Value mode, it will prioritise users similar to your high-value clientele. And if certain demographics truly don’t convert for you, test the demographic exclusion beta by working with your Google rep.

However, a word of caution: don’t over-restrict the campaign. PMax thrives on freedom to test and find conversions in unexpected places. Use exclusions to cut obvious waste (like irrelevant searches or locations you don’t serve) but avoid narrowing the targeting so much that the algorithm can’t explore and learn.

3. Asset Quality and Variety
Creative assets are the fuel for Performance Max. You provide text, images, and optionally videos; Google’s AI then mixes and matches them to create ads across formats. To maximise results, supply a diverse set of high-quality assets.

Include the maximum allowed headlines and descriptions, covering different value propositions and keywords relevant to your audience. Use a mix of product shots and lifestyle images. Don’t re-upload the same Merchant Center images as standalone assets – instead, add lifestyle or branding visuals.

Ensure videos (even short slideshows) are included, as YouTube inventory is a big part of PMax reach. Google will favour the best-performing assets, but it needs a range of options to optimise. Use the asset reporting dashboard to replace underperformers and keep branding consistent across all ad formats.

4. Audience Signals and Segmentation
PMax auto-targets based on data and intent, but audience signals help guide the algorithm. Upload first-party data segments – such as converters or high-LTV customers – and use Google’s custom segments (e.g. based on keyword searches or in-market behaviour).

These signals don’t limit reach but help Google learn faster. For best results, organise asset groups around specific personas or product categories, each with its own creatives and signals. For example, a B2B SaaS brand might separate campaigns for small businesses and enterprise IT managers.

Audience signals act as starting points; Google will expand beyond them. Giving the algorithm a strong foundation speeds up learning and improves early performance.

5. Continuous Monitoring and Optimisation
Set-and-forget doesn’t work. Schedule regular check-ins on campaign performance. Use the new search term insights to exclude irrelevant queries. Monitor conversions and ROAS – if asset groups are underperforming, adjust your segmentation or targets.

In GA4, build a segment for PMax traffic to understand on-site behaviour. Which pages are they landing on? Are they engaging? This can inform future creative and site changes. Reallocate budgets if needed: PMax will spend what it’s given, so ensure high-performing campaigns get appropriate budget.

Consider A/B testing PMax against a standard Search campaign to gauge incremental value (even if it’s not a perfect 1:1 comparison). Finally, review Google’s optimisation suggestions – but treat them as recommendations, not gospel. Your data and KPIs should always come first.

Final Thought:
By mastering these strategies – from embracing new features to fine-tuning your inputs – you can turn Performance Max into a powerful revenue driver. 2025 is all about balancing automation with informed oversight. Let Google’s AI handle the mechanics while you focus on creative direction, smart strategy, and data-informed decisions.

You may also like to read

Fill out the information below