Google Performance Max for Ecommerce: What It Is, How It Works, and When to Use It

In 2022, Google retired Smart Shopping campaigns and replaced them with Performance Max. If you've been running Google Shopping for more than a few years, you'll remember the transition — and probably the frustration that came with it. Less control, less transparency, more "trust the algorithm."

Three years on, Performance Max (PMax) has matured. It can deliver excellent results for ecommerce stores when set up correctly — and disappointing results when it isn't. This guide explains how it actually works, what the trade-offs are, and how to get the most from it.

What Is Performance Max?

Performance Max is a Google Ads campaign type that runs ads across all of Google's inventory from a single campaign:

• Google Search (text ads)

• Google Shopping (product listing ads)

• YouTube (video ads)

• Display Network (banner ads on websites)

• Gmail (sponsored inbox messages)

• Google Discover (content feed ads)

You provide Google with assets — product feeds, images, headlines, descriptions, videos — and Google's machine learning decides where and when to show ads to maximise your conversion goal.

For ecommerce stores, PMax primarily uses your Google Merchant Center product feed to drive Shopping ads. The other channels (YouTube, Display, Gmail) are supplementary.

How Performance Max Actually Works

PMax uses Google's Smart Bidding and machine learning to optimise towards a goal you set — typically "maximise conversion value" or "target ROAS" (return on ad spend).

Google collects signals from:

• Your conversion data (purchases, values)

• Your audience lists (remarketing lists, customer lists)

• The asset groups you create (product categories, creative assets)

• Real-time user behaviour signals (search intent, time of day, device, location)

The more conversion data Google has, the better PMax performs. A new store with few conversions will see poor Performance Max results. A store processing 50+ conversions per month will see the algorithm make smarter decisions. This is the single most important thing to understand about PMax.

Performance Max vs Standard Shopping: What Changed

If you ran Smart Shopping campaigns before PMax, the experience is similar but expanded. Key differences from Standard Shopping:

What you lose:

• Keyword-level control (you can't bid on or exclude specific search terms in the traditional way)

• Placement-level control (you can't exclude individual websites from Display)

• Full search term transparency (the search terms report is less detailed)

What you gain:

• Access to all Google inventory from one campaign

• Audience signals that influence targeting across all channels

• Automated bidding that factors in more signals than manual or smart bidding in Shopping alone

For most ecommerce stores, the shift to PMax is a net positive if conversion tracking is set up correctly. The algorithm genuinely does optimise well once it has data to work with.

Setting Up Performance Max Correctly

The setup choices you make at the start have an outsized impact on results.

1. Connect Your Merchant Center Feed

For ecommerce, a well-optimised Merchant Center feed is the foundation. PMax uses your product titles, descriptions, images, and attributes to determine when and where to show Shopping ads. If your feed is poorly optimised, PMax will underperform regardless of your budget or bidding strategy.

Prioritise these feed attributes:

• Product title: Include primary keyword, brand, colour, size where relevant

• Product description: Clear, accurate, keyword-relevant (Google uses this for search matching)

• Product images: High quality, clean background, no text overlays

• GTIN/MPN: Correct product identifiers help Google match your products to relevant searches

• Product type / custom labels: Useful for organising asset groups

2. Create Thoughtful Asset Groups

An asset group is where you add your creative assets (headlines, descriptions, images, videos) and link them to a subset of products. Think of asset groups like ad groups in Search campaigns.

For ecommerce, organise asset groups by product category rather than putting all products in one group. This lets you:

• Write copy relevant to each category

• Assign audience signals appropriate to each category

• Review performance by product type

Example structure:

• Asset Group 1: Running Shoes → audience signal: people who've viewed running shoe pages

• Asset Group 2: Trail Shoes → audience signal: outdoor sports audiences

• Asset Group 3: Kids' Shoes → audience signal: parents, family audiences

3. Add Audience Signals

Audience signals are not targeting — Google can show ads to people outside your signals. But signals tell Google where to start looking, which is particularly important in the early learning phase when the algorithm has limited data.

Priority audience signals to add:

• Your customer list (upload as a customer match list)

• Website visitors (from your remarketing tag)

• Cart abandoners (from your remarketing audience)

• Custom intent audiences (people who've searched for your top keywords)

4. Choose the Right Bidding Strategy

For established ecommerce stores with sufficient conversion data:

• Target ROAS is the recommended bidding strategy. Set your initial ROAS target conservatively — within 20–30% of your current actual ROAS — and adjust upward gradually as the campaign learns.

For newer stores or campaigns in the learning phase:

• Maximise conversion value (no ROAS target) lets the algorithm optimise without the constraint of a ROAS target you haven't established yet.

Avoid setting an aggressive ROAS target from day one. It will restrict delivery significantly and the campaign may never get out of the learning phase.

5. Set Brand Exclusions

This is critical and often missed. By default, PMax will bid on your brand name and cannabilise traffic from direct searches and organic results. To prevent this, use the Brand Exclusions feature (under campaign settings) to exclude your brand name from triggering PMax ads. Let direct brand traffic come through naturally.

What Good Performance Max Management Looks Like

PMax doesn't mean "set and forget." Active management matters.

Weekly checks:

• Impressions and impression share by asset group

• Conversion volume and ROAS trends

• Budget pacing

Monthly checks:

• Search category insights (the closest thing to a search terms report)

• Asset performance ratings (Produx, Good, Best — Google scores each headline and description)

• Audience insights (what types of users are converting)

When to make changes:

• If ROAS drops significantly over 2+ weeks, check for feed issues first, then audience changes

• If certain asset groups are generating few impressions, review the product feed for those products and the relevance of the creative assets

• Avoid making frequent budget changes — Google's algorithm needs stability to learn

Common Performance Max Mistakes

Putting all products in one asset group. The algorithm can't differentiate messaging between product categories. Create one asset group per category minimum.

Launching with a high ROAS target. If you set a target of 600% ROAS but your historical ROAS is 350%, the campaign will struggle to spend and never gather enough data. Start with your actual ROAS, not your goal.

Not uploading a customer list. Customer lists are one of the strongest audience signals you can give PMax. If you haven't uploaded your customer database as a customer match audience, do it before launching.

Not excluding brand keywords. Without brand exclusions, PMax will spend budget showing ads to people who already know you and were going to visit anyway.

Pausing the campaign during the learning phase. PMax needs 2–4 weeks and a minimum of 30–50 conversions to exit the learning phase. Pausing mid-learning resets the process.

Is Performance Max Right for Your Store?

PMax works well when:

• You have 30+ monthly conversions (ideally 50+) for the algorithm to learn from

• Your Merchant Center feed is well-optimised

• You have creative assets ready (images, at least one video)

• You're willing to manage it actively for the first 4–6 weeks

PMax may not be the best starting point if:

• You're a new store with limited conversion data

• Your product catalogue is small (under 20 products)

• You need precise control over search term targeting

In those cases, Standard Shopping is still a legitimate and often better option — you get more control and clearer visibility into what's working.

Performance Max and Standard Shopping Together

Contrary to popular belief, you can run both Performance Max and Standard Shopping campaigns simultaneously. Standard Shopping will take priority for Shopping inventory when both are running. This lets you retain control over Shopping placements while using PMax for the expanded Google inventory (Display, YouTube, Discover).

This hybrid approach works particularly well for stores that want Standard Shopping precision for their core products while using PMax's broader reach for prospecting and top-of-funnel.

Google's ad landscape changes quickly, and Performance Max will continue to evolve. If you'd like help structuring your Google campaigns for better efficiency and ROAS, our Google Shopping service includes PMax campaign builds and ongoing management.