The Elite Brands PMax Optimisation Framework for eCommerce Success

Performance Max is not a black box.

It feels like one. I get it. When I was scaling my own stores, the idea of handing the keys to a Google algorithm felt like a massive gamble. We see the same hesitation from founders every week. They see PMax as a machine that eats budget and gives unpredictable results.

They are not entirely wrong. If you set it up poorly, that’s exactly what it does.

But when you use a structured approach, PMax becomes the most powerful scaling tool in the Google Ads ecosystem. We’ve seen it time and again across the eCommerce accounts we manage. The difference between a campaign that burns cash and one that prints it is having a repeatable framework.

At Elite Brands, we developed a three-phase framework for exactly this purpose. It turns PMax from a gamble into a system.

1. Phase 1: Strategic Setup. We build the right foundation with a perfect feed and clear signals.
2. Phase 2: Iteration and Refinement. We actively manage and test assets and audiences.
3. Phase 3: Analysis and Scaling. We use data to make smart budget decisions and drive growth.

This is the system we use to manage millions in ad spend. It works because it imposes control on a system that seems to have none.

Phase 1: Strategic setup and feed optimisation for Performance Max

You cannot build a strong house on a weak foundation. For PMax, your product feed is the foundation. Everything the algorithm does, from who it targets to what ad it shows, starts with the data in your Google Merchant centre feed.

When we take on a new client, the first place we look is the feed. I’d say 9 out of 10 have serious, ROAS-killing errors hiding in plain sight. Getting this right is not optional.

The critical role of product feed quality

The PMax algorithm doesn’t “see” your products. It reads your data. If that data is vague or incomplete, your performance will be too.

We focus on optimising the core components first. Product titles need to be descriptive and keyword-rich, following a clear structure like “Brand - Product Type - Key Feature - Colour”. This is far better than a generic title like “Blue T-shirt”. The algorithm needs the detail to match products to high-intent search queries.

Descriptions should be loaded with benefits and features. Think about what a customer would search for and make sure those terms are included.

Custom labels are one of the most underused tools. We use them to segment products for reporting and campaign structure. You can create labels for “bestsellers”, “high-margin”, “seasonal”, or “clearance”. This allows you to build campaigns around specific business goals, not just product categories.

Finally, ensure your data is complete. High-resolution images, accurate pricing, stock levels, and correct GTINs are non-negotiable. Google rewards complete, high-quality data with better visibility.

Foundational campaign settings and audience signals

With a clean feed, you can build the campaign. The first step is precise conversion tracking. We always implement enhanced conversions, which gives Google more accurate data to optimise against, especially with privacy changes impacting cookies. If you sell products at different price points, setting up conversion value rules is also critical to help the algorithm prioritise higher-value sales.

Next are audience signals. This is where you give the algorithm its starting directions. It’s a hint, not a restrictive targeting setting. We start by uploading customer lists, specifically past purchasers and high-LTV segments.

We also build custom segments based on search activity and competitor URLs. This tells Google to look for people who behave like customers of your direct competitors. These signals give PMax a strong starting point, shortening the learning phase and improving initial results. The complexity here is why many brands seek expert Google Ads management.

The last piece of the setup is protection. We implement account-level negative keyword lists to prevent wasted spend on irrelevant terms. We also apply brand exclusions if we want to isolate brand traffic to a dedicated Search campaign. This basic housekeeping saves thousands of dollars.

Phase 2: Asset group iteration and signal refinement for PMax

PMax is not a “set and forget” campaign type. Phase 1 gives you a solid launchpad. Phase 2 is where the real, ongoing work happens. This is about continuous testing and refinement based on performance data.

The goal is to systematically improve the inputs you give the algorithm. Better assets and sharper signals lead to better outputs.

Crafting high-performing asset groups

We never dump all products into a single asset group. We create thematic groups based on product category, brand, or promotion. For example, a fashion brand might have separate asset groups for “Women’s Dresses”, “Men’s Outerwear”, and “Summer Sale”. This allows us to tailor creative and messaging specifically to the products within that group.

Within each asset group, we run a constant cycle of A/B testing. We test different headlines, comparing benefit-led copy against feature-focused copy. We test lifestyle imagery against clean product-on-white shots.

Video is a huge component. We’ve found that short, 15-30 second videos designed for YouTube Shorts and Display placements significantly lift performance. You don’t need a Hollywood budget. Simple, product-focused videos work well.

Pay attention to Google’s “Ad Strength” indicator, but don’t treat it as the absolute truth. We’ve seen asset groups with an “Average” rating outperform ones with an “Excellent” rating. The real source of truth is your ROAS and CPA data.

Refining audience signals for PMax

The audience signals you start with are not the ones you finish with. They need to evolve.

We have a process for regularly updating customer match lists. Every 30-60 days, we upload fresh lists of recent purchasers and new VIP customers. This keeps the algorithm fed with the most current data on who your best customers are.

We also iterate on custom segments. We use Google Trends and search term insight reports to identify new keywords and competitor brands to add to our signals. The market is always changing, and your signals need to reflect that.

This is also where we bring in data from other platforms. The insights we get from our Meta Ads management are invaluable. If a certain message or audience is performing exceptionally well on Facebook, we build a corresponding custom segment or test that messaging in our PMax asset groups. This creates a powerful feedback loop across your entire marketing mix.

Phase 3: Performance analysis and budget scaling for PMax success

Once your campaigns are running and you have a solid testing process, the focus shifts to analysis and scaling. This is how you turn a profitable campaign into a major growth engine for your brand. It requires a disciplined approach to reading the data and making strategic decisions.

Simply increasing the budget on a campaign that “looks good” is a recipe for wasted spend. We dig deeper to understand what is working and why, before we invest more.

Deep diving into PMax performance data

We monitor a core set of metrics beyond the top-line ROAS. We analyse conversion value, cost per acquisition, and impression share to get a complete picture of performance. A high ROAS is great, but if your impression share is only 10%, there’s a huge amount of untapped potential.

The Insights tab within a PMax campaign is your best friend. It provides data on which audience segments are driving conversions and what search categories are triggering your ads. This used to be a complete black box, but Google has improved visibility here. You can use these insights to refine your audience signals and asset group creative. For more detail, Google’s own documentation on evaluating Performance Max performance is a good starting point.

We also drill down to the product level. By connecting Google Ads to Merchant centre, you can see which specific SKUs are performing well and which are dragging down your ROAS. We often pause underperforming products within a campaign to force the budget towards the winners. This simple optimisation can lift a campaign’s overall return significantly.

Strategic budget scaling and optimisation

Knowing when and how to scale budget is critical. We never make large, sudden changes. The algorithm needs time to adjust.

Our rule is to wait until a campaign is out of its learning phase and has shown stable performance for at least two weeks. Then, we increase the budget in small increments, typically no more than 20% at a time. We wait a few days, analyse the impact on ROAS and CPA, and then decide if we can scale again. This methodical approach prevents the performance shocks that come from doubling a budget overnight.

If you hit a performance plateau, it’s time to troubleshoot. Is your ROAS dropping as you increase spend? This could be a sign of audience saturation. It might be time to introduce new audience signals or refresh your creative assets. Seeing our clients overcome these plateaus and achieve consistent growth is what drives us. You can see some of our results to understand the impact of this structured approach.

Integrating Performance Max with other Google Ads channels

Performance Max doesn’t exist in a vacuum. It interacts with every other campaign in your Google Ads account. Understanding this ecosystem is crucial for maximising your total account performance and avoiding costly mistakes.

A common mistake I see is brands running PMax alongside other campaigns without a clear strategy. This often leads to internal competition, where your own campaigns bid against each other, driving up your costs.

PMax has a complex relationship with Search campaigns. For most search queries, if you have an eligible ad in both a PMax campaign and a Search campaign, PMax will typically be prioritised. However, exact match keywords in your Search campaigns will still take priority over PMax. This is why we almost always run a dedicated Brand Search campaign. It protects our clients’ most valuable keywords and gives us full control over that traffic.

When it comes to Shopping, PMax has priority over Standard Shopping campaigns for the same products. For most eCommerce brands, this means PMax effectively replaces Standard Shopping. However, there can be niche cases where a Standard Shopping campaign with specific settings might be useful.

PMax also serves ads across the Display and YouTube networks. If you are running separate campaigns for these channels, you need to be mindful of audience overlap. A unified strategy ensures your ads are complementary, not cannibalistic. Thinking about this a level deeper is a core part of our process. A holistic view of the account ensures every dollar is working as hard as possible.

Unlocking eCommerce growth with PMax best practices

Managing Performance Max effectively isn’t about finding one secret hack. It’s about implementing a disciplined, continuous system.

Our three-phase framework provides that system.

1. Setup: A perfectly optimised feed and smart initial signals give you the best possible start.
2. Iteration: Constant testing of assets and refinement of audiences keeps performance improving.
3. Scaling: A data-led approach to analysis and budget changes drives sustainable growth.

This isn’t a one-off project. It’s an ongoing process of optimisation. The brands that win with PMax are the ones that treat it as a core operational rhythm within their marketing. They are constantly feeding the machine better data, better creative, and clearer signals.

The result is a scalable, predictable engine for acquiring new customers. It moves you from gambling on an algorithm to directing it with intent. This is the key to unlocking consistent growth for your eCommerce brand.

If you’re running PMax campaigns and feel like you’re stuck in the “black box”, a structured approach can make all the difference. We offer a free Google audit to show you exactly where the opportunities are in your account.

Getting this right can fundamentally change the growth trajectory of your business.