Dynamically - AI Marketing Agency
PPC

Performance Max Campaigns: A Complete Guide for Ecommerce

Szymon10 min read
Performance Max Campaigns: A Complete Guide for Ecommerce

Performance Max has become Google's flagship campaign type, and for ecommerce advertisers it represents both a tremendous opportunity and a significant shift in how campaigns are managed. Since replacing Smart Shopping in 2022, Performance Max has evolved considerably – yet many advertisers still struggle to unlock its full potential, or worse, let it run unchecked and wonder where their budget went.

This guide covers everything ecommerce businesses need to know about Performance Max campaigns in 2026: how they work under the bonnet, how to set them up properly, and – critically – how to optimise them when Google's automation doesn't get it right on its own.

What Is Performance Max?

Performance Max (PMax) is a goal-based campaign type that uses Google's AI to serve ads across all of Google's advertising channels from a single campaign. That includes:

  • Google Search (including text ads)
  • Google Shopping
  • YouTube
  • Gmail
  • Google Display Network
  • Discover feed
  • Google Maps

Rather than creating separate campaigns for each channel, you provide Google with your creative assets (images, videos, headlines, descriptions), your product feed, your conversion goals, and your budget. Google's algorithm then decides which combination of assets to show, on which channel, to which user, and at what bid – all in real time.

For ecommerce specifically, PMax campaigns pull product data from your Google Merchant Centre feed and can serve Shopping ads, dynamic display ads, and video ads featuring your products automatically.

How Performance Max Actually Works

The Algorithm's Decision-Making

At its core, PMax uses Google's machine learning to optimise towards your specified conversion goal – typically purchases or revenue for ecommerce. The system considers hundreds of signals in real time:

  • The user's search query (if on Search or Shopping)
  • Their browsing history and in-market signals
  • Time of day, device, and location
  • Which creative assets are most likely to drive a conversion
  • Historical conversion data from your account

The algorithm needs data to learn. Google typically recommends a minimum of 30 conversions within a 30-day period for PMax to optimise effectively. Campaigns with fewer conversions will struggle to exit the learning phase, leading to erratic performance and wasted spend.

Asset Groups: Your Building Blocks

Asset groups are the core structural unit of a Performance Max campaign. Each asset group contains:

  • Final URL: The landing page users will reach
  • Images: Up to 20 images in various aspect ratios (landscape 1.91:1, square 1:1, portrait 4:5)
  • Logos: Up to 5 logos
  • Videos: Up to 5 videos (if you don't provide video, Google will auto-generate one – and it's usually terrible)
  • Headlines: Up to 15 headlines (30 characters max each)
  • Long headlines: Up to 5 (90 characters max each)
  • Descriptions: Up to 5 (90 characters max each), plus 1 short description (60 characters)
  • Call to action: Automated or manual selection
  • Audience signals: Suggested audiences for the algorithm to start with
  • Listing groups: Which products from your feed to include

For ecommerce, you'll typically create multiple asset groups – one per product category or brand – each with tailored creative and a relevant subset of your product catalogue.

Audience Signals vs Audience Targeting

This is one of the most misunderstood aspects of PMax. Audience signals are suggestions, not restrictions. When you add a custom segment, in-market audience, or customer list as an audience signal, you're telling Google: "Start by looking at people like this, but feel free to expand beyond them if you find better opportunities."

This means PMax will always reach beyond your specified audiences. You cannot lock it down to a specific audience the way you can with a standard Display or Video campaign. For some advertisers this is a strength – Google often finds high-converting users you wouldn't have thought to target. For others, it means budget leaking into low-quality placements.

Setting Up PMax for Ecommerce: Step by Step

1. Get Your Feed Right First

This cannot be overstated: your product feed is the single most important factor in PMax performance for ecommerce. The campaign can only be as good as the data it's working with.

Before launching PMax, ensure your Shopping feed is fully optimised:

  • Product titles include relevant keywords (brand, product type, key attributes)
  • GTINs are present and accurate for all applicable products
  • Product images are high-quality, white-background for primary images
  • Pricing is competitive and accurate
  • Product categories are correctly mapped to Google's taxonomy
  • Custom labels are set up for segmentation (margin tiers, best sellers, seasonal items)

2. Structure Your Asset Groups Strategically

Don't dump all your products into a single asset group. Instead, segment by:

  • Product category: Shoes, bags, accessories each get their own asset group with category-specific creative
  • Performance tier: Separate your best sellers from long-tail products – this lets you allocate budget more effectively
  • Margin tier: Products with higher margins can sustain higher CPAs, so they benefit from separate bidding strategies

A typical ecommerce PMax structure might look like:

  • Campaign 1 (High Priority): Best sellers and hero products – higher budget, target ROAS aligned to margin
  • Campaign 2 (Standard): Core catalogue – moderate budget, standard ROAS target
  • Campaign 3 (Catch-all): Remaining products – lower budget, exploratory

3. Provide Strong Audience Signals

The better your audience signals, the faster PMax will learn. Prioritise these signal types:

  • Customer match lists: Upload your existing customer email list. This gives Google the strongest signal about who your buyers actually are.
  • Website visitor lists: Add remarketing audiences, particularly past purchasers and high-intent visitors (add-to-cart, checkout initiated).
  • Custom segments: Create segments based on search terms your customers use and URLs they visit (including competitor sites).
  • In-market audiences: Layer relevant in-market segments for your product category.

4. Create Compelling Creative Assets

PMax runs across visual channels (YouTube, Display, Discover, Gmail), so your creative assets matter significantly. Key tips:

  • Always upload your own video. Google's auto-generated videos are a slideshow of your images with generic transitions. They look cheap and perform poorly. Even a simple 15-second product video will outperform them.
  • Provide all image ratios. If you only supply landscape images, your ads won't serve effectively on mobile-first placements (which is most of them).
  • Write headlines for humans, not algorithms. Your headlines appear in text ads, overlay on images, and as YouTube bumper text. They need to make sense in multiple contexts.
  • Include lifestyle and product images. A mix of white-background product shots and in-context lifestyle imagery tends to outperform either approach alone.

5. Set Realistic Bidding Targets

For ecommerce, you'll typically use one of two bidding strategies:

  • Maximise Conversion Value: Google spends your full budget to generate the highest possible revenue. Use this during the learning phase when you don't yet have enough data for a target ROAS.
  • Target ROAS: You set a target return on ad spend (e.g., 400% means £4 revenue for every £1 spent). Only use this once you have at least 30 conversions in 30 days and a clear understanding of your break-even ROAS.

Setting a target ROAS too aggressively from day one is one of the most common mistakes. It starves the campaign of data, prevents learning, and usually results in poor performance. Start with Maximise Conversion Value, let the campaign accumulate data for 4–6 weeks, then introduce a target ROAS that's 10–20% below your observed average.

Reporting Limitations – and Workarounds

PMax's biggest weakness is transparency. Compared to traditional campaign types, you get significantly less insight into where your budget is going and why. Here's what you're missing, and what you can do about it:

Search Term Visibility

PMax does not show you a full search terms report. You can see "search categories" and "search themes" in the Insights tab, but not the individual queries triggering your ads. This makes it difficult to identify irrelevant traffic or wasted spend.

Workaround: Run a standard Search campaign alongside PMax for your most important keywords. This gives you search term data for those queries and ensures PMax doesn't cannibalise your highest-intent traffic.

Channel Performance Breakdown

You can see high-level placement data (how much spend went to Search vs Display vs YouTube), but the granularity is limited. You can't easily see which specific placements or audiences are driving results.

Workaround: Use Google Ads scripts or third-party tools to pull placement-level data via the API. Several free scripts are available that extract this data into a Google Sheet.

Asset Performance

Google rates your assets as "Low," "Good," or "Best," but doesn't provide detailed metrics for individual assets. You know which headlines are performing well, but not exactly how well.

Workaround: Rotate assets regularly. Replace "Low" performing assets every 2–4 weeks and test new variations. Treat it as a continuous creative testing programme.

Optimisation Tips for Ecommerce PMax

1. Use Negative Keywords (Yes, You Can Now)

As of 2024, Google allows account-level negative keywords that apply to PMax campaigns. Use these aggressively to prevent your ads from showing on irrelevant or brand-cannibalising queries. Common exclusions include:

  • Competitor brand terms (unless you intentionally want to bid on them)
  • "Free," "cheap," "DIY" – terms that attract low-quality traffic
  • Job-related queries ("careers at [brand]," "[product] jobs")
  • Informational queries that don't convert ("[product] review," "how to [product]")

2. Exclude Low-Performing Products

Not every product deserves PMax budget. Use listing group exclusions to remove:

  • Products with no margin
  • Out-of-season items
  • Low-stock products
  • Products with poor imagery or incomplete feed data

3. Use Custom Labels for Segmentation

Custom labels in your Merchant Centre feed are essential for PMax segmentation. Common labelling strategies:

  • custom_label_0: Margin tier (high, medium, low)
  • custom_label_1: Performance tier (best seller, core, long tail)
  • custom_label_2: Seasonality (spring, summer, evergreen)
  • custom_label_3: Price range (premium, mid-range, value)
  • custom_label_4: Promotion status (on sale, clearance, full price)

4. Layer in First-Party Data

The more first-party data you feed PMax, the better it performs. This includes:

  • Customer email lists (segmented by lifetime value if possible)
  • Enhanced conversions (passing hashed customer data back to Google)
  • Offline conversion imports (if you have phone or in-store sales)
  • Consent-mode-compliant tracking (essential for UK/EU compliance)

5. Don't Set and Forget

Despite being "automated," PMax requires regular attention. A sensible optimisation cadence for ecommerce PPC:

  • Daily: Check spend pacing and flag any anomalies
  • Weekly: Review asset performance, search themes, and product-level ROAS
  • Fortnightly: Refresh underperforming creative assets
  • Monthly: Reassess ROAS targets, product segmentation, and budget allocation
  • Quarterly: Full structural review – are your asset groups still aligned with your product range and business goals?

Common PMax Mistakes to Avoid

  1. No product feed optimisation: Running PMax with a basic, unoptimised feed is like entering a race with flat tyres. Fix your feed first.
  2. Single asset group for everything: Lumping all products together gives the algorithm no structure to work with and makes optimisation impossible.
  3. Skipping video: Google will auto-generate poor-quality videos. Always provide your own.
  4. Setting ROAS targets too early: Let the campaign learn before constraining it.
  5. Ignoring cannibalisation: PMax can cannibalise your existing Search and Shopping campaigns if not structured carefully. Run branded Search separately and monitor for overlap.
  6. No negative keywords: Without exclusions, PMax will inevitably spend on irrelevant queries.

Is Performance Max Right for Your Ecommerce Business?

PMax works best when:

  • You have a catalogue of 50+ products with a well-optimised feed
  • You generate at least 30 conversions per month (ideally more)
  • You have strong creative assets (images, video, compelling copy)
  • You have first-party data (customer lists, robust tracking)
  • You're prepared to invest in ongoing optimisation – not just set it up and walk away

If you're a smaller ecommerce business with fewer products or lower conversion volume, you may get better results from standard Shopping campaigns where you have more control over bids and targeting.

At Dynamically, we manage Performance Max campaigns for ecommerce businesses across the UK, from scaling DTC brands to established retailers with thousands of SKUs. Whether you're launching PMax for the first time or struggling to get results from an existing campaign, our team can audit your setup, restructure your campaigns, and optimise your product feed to drive measurable improvement. Get in touch to discuss your ecommerce PPC strategy.

Work with Dynamically

Ready to put these insights into practice?

Our Liverpool-based team works with UK businesses to grow organic search, improve paid media performance and build visibility in AI-powered search. Get a free audit to see exactly where your opportunities are.