Optimizing Performance Max for Online Shops 2026

Google Performance Max has established itself as the most important campaign type in e-commerce: 71% of advertisers already use PMax (Fluency), achieving on average 27% more conversions at comparable CPA (Google Ads Help). But between an automated setup and an optimized campaign lies the difference between a mediocre ROAS of 2.8x and professional 4.1x or more (WordStream-LocaliQ). In this guide, we show online shop operators how to systematically optimize their Performance Max campaigns for maximum return.

What Is Performance Max and Why Does It Matter?

Performance Max (PMax) is Google's AI-powered campaign type that automatically serves ads across all Google channels: Search, Shopping, Display, YouTube, Gmail, Discover and Maps. Instead of managing individual campaigns for each channel, shop operators provide assets (texts, images, videos) and goals - the AI handles ad delivery and bid optimization. The promise: Google's machine learning algorithms find the right person at the right time on the right channel - fully automated.

Market penetration speaks for itself: 19.9% of total Google Ads spend now goes to PMax, a 19% increase year-over-year (Fluency). In the UK, the adoption rate is as high as 84% (Fluency). Globally, over 60% of advertisers use AI-powered campaign formats like PMax (Google). For Google Ads in e-commerce, Performance Max is no longer optional - it is the standard. Google itself generated $81 billion in ad revenue in 2025 (DemandSage), with Shopping ads accounting for 60% of Google retail ad spend (uproas.io) and approximately 36% of all online purchases influenced by Google Search or Shopping Ads (uproas.io).

The critical insight: Feed-based ads account for 74-97% of PMax spend (smec). This means the quality of your product feed in Google Merchant Center is the biggest lever for campaign success. A campaign is only as good as the data that fuels it. The global digital ad market is projected to reach $850 billion in 2026 (eMarketer), with Google controlling 27% of all digital ad budgets (eMarketer) and 83% of global search ad revenue (uproas.io), remaining the dominant platform.

ROAS Benchmarks: Where Does Your Shop Stand?

Before you optimize, you need a reference point. The current ROAS benchmarks by campaign type show what is realistically achievable:

Notably, there is a correlation between budget and ROAS: Shops with over $50,000 monthly ad spend achieve 5.2x ROAS, while smaller budgets under $2,000 only reach 2.8x (WordStream-LocaliQ). The reason: larger budgets provide more conversion data that Google's AI uses for better optimization. The cross-platform e-commerce ROAS stands at 2.87:1 (Upcounting), showing that Google Ads outperforms social ad platforms on average.

Product Feed Optimization as the Foundation

Since 90% of PMax spend (median) flows into feed-based ads (smec), the product feed is the most important success factor. A poorly optimized feed means wasted budget - regardless of how well the rest of the campaign is configured. Since the Google Merchant API migration, new technical requirements also apply to feed transmission.

A clean product feed starts with data quality in the shop system. Those who already maintain complete product attributes in their Shopware backend or WooCommerce have a decisive advantage in feed optimization. For Shopware shops, a direct connection via the Merchant API is recommended over manual CSV uploads.

Particularly important for e-commerce: Custom Labels enable strategic segmentation of the product catalog. By tagging products by margin groups (high, medium, low), bestseller status or seasonal relevance, you can create targeted asset groups with different ROAS targets within PMax. A fashion shop could run summer sale items in a separate group with lower target ROAS, while new arrivals with higher margins are advertised more aggressively. This approach allows a differentiated pricing strategy even within an automated campaign.

Asset Groups and Creative Strategy

Beyond the product feed, asset groups are the second major optimization lever. Google uses the provided texts, images and videos to automatically create ad combinations for different channels. The quality and variety of assets determines how effectively the AI can serve ads.

Videos have a measurable impact: Campaigns with at least one video achieve 12% more conversions (Google Ads Help). Those who provide videos in all three orientations (horizontal, vertical, square) achieve 20% more YouTube conversions than with horizontal videos alone (Google Ads Help). Without custom videos, Google creates auto-generated clips from images and text - but their quality is limited. Professionally produced 15-30 second videos showing the product in action typically outperform auto-generated variants significantly.

• At least 5 headline texts (max 30 characters) with different USPs
• At least 5 long headlines (max 90 characters) with product benefits and CTAs
• At least 5 images in various formats (1200x628, 1200x1200, 960x1200)
• At least 1 video (ideally 3: horizontal, vertical, square)
• Sitelink extensions with relevant landing pages
• At least 2 description texts (max 90 characters)

Bidding Strategies and Conversion Data

The choice of bidding strategy directly impacts performance. 95% of PMax campaigns use "Maximize Conversion Value" as their strategy (smec) - and for good reason: this strategy optimizes for the highest revenue, not just the number of conversions. In e-commerce, this is a crucial difference - a shop does not want ten orders at $10 each, but ideally high-value carts with correspondingly higher contribution margins.

The data shows a clear difference: Campaigns with "Max Conv. Value" achieve a median conversion rate of 2.22%, while "Max Conversions" only reaches 1.98% (smec). For online shops optimizing for revenue rather than leads, the difference is significant.

Critical for success is the conversion data foundation: Google recommends at least 30 conversions per month for reliable performance, with 60 or more being optimal (smec). Campaigns below 30 conversions show extreme variance from -100% to +400% against target ROAS (smec). Those with too few conversions should initially optimize for micro-conversions (add to cart, checkout initiation) and later switch to purchase.

PMax and Standard Shopping: The Hybrid Approach

Since October 2024, Ad Rank determines which campaign wins the auction - PMax is no longer automatically prioritized (Google). This opens new strategic possibilities: The hybrid approach combining PMax and Standard Shopping has become the standard in 2026 (smec/Search Engine Land). For shop operators, this change means they have genuine control for the first time over which campaign is served for which search queries - a paradigm shift compared to the previous PMax prioritization.

The logic: Standard Shopping offers the highest return at 5.2x ROAS with full control over bids and keywords (WordStream-LocaliQ). Performance Max complements with its reach across Display, YouTube and Discover, capturing users in earlier stages of the customer journey. The optimal structure involves 3-7 campaigns per account (smec).

Google's own recommendation for 2026 is the "Power Pack" combination: Performance Max + Demand Gen + AI Max for Search (Google Marketing Live 2025). Demand Gen alone has increased its spend by +192% (Fluency), reflecting growing interest in upper-funnel channels. For e-commerce operators, this means PMax alone is not enough - a well-designed campaign architecture across multiple campaign types delivers the best results.

Negative Keywords and Avoiding Wasted Budget

One of the biggest criticisms of Performance Max is the lack of search query transparency. Google has expanded negative keyword capabilities for PMax, but control remains limited. Nevertheless, there are effective strategies to minimize budget waste.

Adding irrelevant terms like "free" or "used" as negative keywords can reduce costs by up to 15% (smec). Account-wide negative keyword lists for PMax campaigns can be set up through Google Ads support. Additionally, URL exclusions help prevent PMax from driving traffic to irrelevant pages like your imprint or privacy policy.

1. Review search terms report regularly (under "Insights" in Google Ads)
2. Configure brand exclusions to exclude competitor brand names
3. Set URL exclusions for non-commercial pages (blog, imprint, terms)
4. Add placement exclusions for low-quality websites in the Display network
5. Request account-wide negatives for PMax via Google support

Another lever is the feed-only strategy: PMax campaigns without additional assets (product feed only) deliver 20-30% higher ROAS at lower volume (Ads Kong). This approach is particularly suitable for shops with limited budgets that prioritize efficiency over reach. No text, image or video assets are provided - the campaign uses only the product feed and primarily serves Shopping ads. The downside: reach via YouTube and Display is eliminated, which limits new customer acquisition.

Conversion Tracking as a Prerequisite

Without proper conversion tracking, any PMax optimization is flying blind. The AI needs precise signals to correctly manage bids. Especially in the context of iOS privacy updates and cookie restrictions, server-side tracking is becoming mandatory for e-commerce. When the AI receives incorrect or incomplete conversion data, it optimizes for the wrong audiences - with direct impact on ROAS.

For online shops, the combination of Enhanced Conversions and server-side tracking is recommended. Enhanced Conversions compensate for data loss from cookie blocking by transmitting hashed customer data (email, phone) to Google. Combined with properly implemented consent management and privacy-compliant tracking, this creates a solid data foundation for AI optimization.

An often underestimated aspect is the consent rate: The more users agree to tracking, the more data the AI receives for optimization. Shops with high consent rates (above 80%) have a structural advantage in PMax performance compared to shops where only 40-50% of users consent. Optimizing the consent banner is therefore indirectly a measure to improve campaign performance. Cookieless tracking strategies can further reduce this dependency.

Mobile Performance and Vertical Video

Mobile devices play a central role in Performance Max. Current benchmarks paint a nuanced picture: Mobile CTR is 4.2% versus 3.5% on desktop (MetricNexus), with simultaneously 20% lower CPC (MetricNexus). However, desktop still converts better at 4.3% compared to mobile at 3.5% (MetricNexus).

The most important mobile trend in 2026: Vertical video formats achieve 40% lower CPMs than horizontal formats (Ads Kong). Since YouTube Shorts and Discover are primarily used on mobile, vertical videos are a cost-effective way to increase reach. As conversion rate benchmarks show, mobile conversion in e-commerce generally trails desktop - making an optimized mobile checkout all the more important.

Real-World Results: What PMax Can Deliver

The theory is compelling - but what do real results look like? Several documented case studies demonstrate the potential of optimized PMax campaigns:

Particularly interesting is the "zombie SKU" optimization approach (Ads Kong): Through targeted asset groups for neglected products, one e-commerce company generated +18% revenue and identified 50 new profitable SKUs that previously received barely any impressions. For shops with large catalogs, this is a significant lever to unlock the full potential of the product range.

Data-driven A/B testing also plays a role in PMax optimization: Different asset groups, audience signals and bidding strategies should be systematically tested against each other to find the optimal configuration for your shop. The case studies are particularly promising for new customer acquisition: Nykaa (Beauty) increased new customer acquisitions by +50% (Think with Google), while Hairstory achieved a fivefold increase in new customer conversions (Google). Separating existing customers from new customers in audience signals allows the AI to optimize specifically for new customer acquisition.

Step-by-Step PMax Optimization

Based on benchmarks and best practices, a clear optimization sequence emerges for online shops. The order is deliberate: each step builds on the previous one. Without proper tracking (step 1), the data foundation for everything else is missing.

1. Verify conversion tracking: Enable Enhanced Conversions, validate conversion values, implement server-side tracking
2. Optimize product feed: Complete titles, images, attributes and custom labels in Google Merchant Center
3. Build asset groups: Provide at least 15 images, 5 headlines, 5 descriptions and 1 video per group
4. Choose bidding strategy: "Maximize Conversion Value" with realistic target ROAS (plan for 95-116% achievement rate)
5. Set audience signals: Define customer lists, website visitors and in-market audiences as signals
6. Configure negative keywords: Set up irrelevant terms via support and configure URL exclusions
7. Build hybrid structure: Standard Shopping for top performers, PMax for scaling and new audiences
8. Allow learning phase: Plan at least 6-8 weeks for stable results (smec)

Frequently Asked Questions