Performance Max (PMax) is Google's AI-driven campaign type that serves ads across all Google properties: Search, Shopping, Display, YouTube, Gmail, and Discover.
For ecommerce, it's become the default campaign type—replacing Smart Shopping and expanding reach significantly.
Here's how to set it up correctly and optimize for performance.
What Is Performance Max?
Performance Max uses Google's machine learning to:
- Automatically bid across all Google inventory
- Find converting audiences using your signals
- Test creative combinations dynamically
- Optimize toward your conversion goals
Key difference from Standard Shopping: PMax includes all channels (not just Shopping) and uses automation for targeting, bidding, and creative optimization.
When to Use Performance Max
Good Fit for PMax
- Established conversion tracking: You need reliable data for the algorithm
- Product catalog (feed): Required for Shopping placements
- Sufficient budget: $50+/day per campaign minimum
- Broad product appeal: Products that work across audience types
- Variety of assets: Images, videos, headlines to test
Consider Standard Campaigns Instead
- Brand-new accounts: Limited conversion history for optimization
- Very low budget: Under $30/day per campaign
- Specific audience targeting needs: PMax has limited audience controls
- Detailed reporting requirements: PMax reporting is limited
- Search-only strategy: Standard Search gives more control
Setting Up Performance Max
Step 1: Prerequisites
Before creating a PMax campaign:
1. Conversion tracking: Ensure purchase conversions are tracking properly. Verify in Google Ads > Tools > Conversions.
2. Product feed: Set up in Google Merchant Center. Feed quality directly impacts Shopping performance.
3. Asset collection: Gather images, logos, videos, headlines, and descriptions.
Step 2: Campaign Creation
- In Google Ads, click "+ New Campaign"
- Select "Sales" as campaign objective
- Choose "Performance Max" campaign type
- Link your Merchant Center account
Step 3: Budget & Bidding
Budget:
- Set daily budget (recommend $50-100+ minimum)
- PMax needs budget to explore and learn
Bidding Strategy:
| Strategy | When to Use |
|---|---|
| Maximize conversions | New campaigns, building volume |
| Maximize conversion value | Varied product prices, want revenue focus |
| Target ROAS | Established campaigns with stable performance |
| Target CPA | Consistent product pricing |
Recommendation: Start with "Maximize conversion value" for 2-4 weeks, then add Target ROAS once you have baseline data.
Step 4: Asset Groups
Asset groups are collections of creative assets organized by theme (product category, audience, etc.).
Asset Group Structure Options:
Option A: Single Asset Group
- Best for: Small catalogs, similar products
- Pros: Simple, all budget in one place
- Cons: Less granular control and reporting
Option B: Multiple Asset Groups by Category
- Best for: Diverse product catalog
- Pros: Tailored messaging per category, better reporting
- Cons: More setup, budget fragmentation
Example for apparel brand:
- Asset Group 1: Men's Clothing
- Asset Group 2: Women's Clothing
- Asset Group 3: Accessories
Step 5: Asset Requirements
Each asset group needs:
Text Assets:
- Headlines (3-5): Max 30 characters each
- Long headlines (1-5): Max 90 characters each
- Descriptions (2-4): Max 90 characters each
- Business name: Your brand name
Image Assets:
- Landscape images (1-20): 1.91:1 ratio, 1200x628 recommended
- Square images (1-20): 1:1 ratio, 1200x1200 recommended
- Portrait images (optional): 4:5 ratio
- Logo: Square and landscape versions
Video Assets (Recommended):
- YouTube videos: Landscape, square, or vertical
- If you don't provide video, Google will auto-generate (usually lower quality)
Best Practice: Provide diverse assets. PMax tests combinations—more variety means more testing opportunities.
Step 6: Audience Signals
Audience signals help Google find the right users faster. They're suggestions, not hard targeting.
Signal Types:
Custom segments:
- People who searched for specific terms
- People who visited certain websites
- People who use certain apps
Your data:
- Customer lists (purchasers, high-value customers)
- Website visitors
- App users
Interests & demographics:
- In-market audiences
- Affinity audiences
- Detailed demographics
Recommendation: Add strong signals:
- Past purchaser list (your best customers)
- High-intent search terms (competitors, product categories)
- In-market audiences for your category
Audience signals accelerate learning but don't limit reach. PMax will expand beyond your signals.
Step 7: Final URL Expansion
Final URL expansion: Allows Google to send traffic to any page on your site (not just URLs you specify).
Recommendation: Keep enabled but exclude:
- Blog pages
- About/contact pages
- Out-of-stock products
- Non-converting pages
Set exclusions in campaign settings > Additional settings > Final URL expansion.
Optimizing Performance Max
Week 1-2: Learning Phase
What to expect:
- Volatile performance
- Higher-than-target CPA/lower ROAS
- Google testing audiences and placements
What to do:
- Don't make major changes
- Monitor for tracking issues
- Ensure assets are approved
Week 2-4: Stabilization
What to expect:
- Performance stabilizing
- Clearer trends emerging
- Budget allocation settling
What to do:
- Review asset performance (which images/headlines winning)
- Check placement insights
- Identify underperforming asset groups
Ongoing Optimization
1. Asset Refresh
Add new assets regularly:
- Replace low-performing images/headlines
- Test new video content
- Refresh seasonal messaging
Google reports asset performance as "Low," "Good," or "Best." Replace "Low" performers.
2. Audience Signal Refinement
- Add new customer lists as they grow
- Test different custom segments
- Remove signals that don't correlate with performance
3. Feed Optimization
Your product feed is critical for Shopping placements:
| Feed Element | Optimization |
|---|---|
| Title | Include key attributes (brand, color, size, material) |
| Description | Detailed, keyword-rich, benefit-focused |
| Images | High-quality, white background, multiple angles |
| Price | Competitive, updated regularly |
| Availability | Accurate stock status |
| Product type | Detailed categorization |
| Custom labels | Segment by margin, bestseller, season |
4. Negative Keywords (Brand)
PMax will bid on your brand terms unless excluded. This can inflate ROAS artificially.
To add brand negatives:
- Contact Google support (account-level exclusions)
- Or use scripts to identify brand queries in Search terms report
Performance Max Reporting
PMax reporting is limited compared to standard campaigns. Here's what you can see:
Available:
- Campaign-level metrics (conversions, ROAS, spend)
- Asset group performance
- Asset performance (Low/Good/Best)
- Insights tab (audience, search trends)
- Placement reports (which properties served ads)
Limited/Unavailable:
- Full search term reports
- Detailed audience breakdowns
- Exact placement URLs
- Device-level optimization
Workarounds:
- Use the Insights tab for directional data
- Check "Listing groups" for product-level performance
- Use scripts for enhanced reporting
- Compare PMax period-over-period, not to other campaigns
Common PMax Mistakes
1. Insufficient Budget
PMax needs enough budget to test across channels. Under $50/day fragments spend and extends learning.
Fix: Consolidate into fewer, better-funded campaigns.
2. Poor Asset Quality
Low-quality images, generic headlines, or no video limits what PMax can test.
Fix: Invest in quality assets. Provide variety.
3. Neglecting Feed Quality
Shopping placements depend on feed data. Poor titles and images = poor Shopping performance.
Fix: Optimize feed titles, descriptions, and images regularly.
4. No Audience Signals
Without signals, PMax takes longer to find converting audiences.
Fix: Add customer lists, search term segments, and in-market audiences.
5. Ignoring Brand Cannibalization
PMax will gladly spend on branded searches—easy conversions that inflate ROAS.
Fix: Exclude brand terms or run separate brand campaigns.
6. Set-and-Forget
PMax is automated but not hands-off. Assets go stale, feeds need updates, performance shifts.
Fix: Review weekly. Refresh assets monthly. Optimize feed quarterly.
PMax vs. Standard Shopping: Which to Use?
| Factor | Performance Max | Standard Shopping |
|---|---|---|
| Reach | All Google properties | Shopping only |
| Control | Limited | High |
| Reporting | Limited | Detailed |
| Automation | Full | Partial |
| Best for | Scale, efficiency | Control, testing |
Most ecommerce brands should:
- Run PMax as primary Shopping campaign
- Keep standard Search for brand and high-intent keywords
- Consider standard Shopping for specific product tests
Performance Expectations
Typical Results
Compared to Smart Shopping (predecessor):
- 10-20% more conversions on average
- Similar or slightly better ROAS
- Broader reach (YouTube, Display, Discover)
- More brand/competitor exposure
Timeline
- Weeks 1-2: Learning phase, volatile metrics
- Weeks 3-4: Stabilization, trends visible
- Months 2+: Optimization opportunities, stable baseline
Red Flags
- CPA 50%+ above target after 4 weeks
- Majority spend on Display/YouTube with poor returns
- Brand searches dominating (check Insights)
- No improvement from asset updates
The Bottom Line
Performance Max is powerful but requires proper setup:
- Quality product feed
- Diverse creative assets
- Strong audience signals
- Sufficient budget
- Ongoing optimization
It's not set-and-forget—but when managed well, PMax can efficiently scale ecommerce advertising across Google's entire ecosystem.
Treat it as a core acquisition channel alongside Meta, and optimize both using blended metrics that measure real business impact.