How to Use Audience Signals in Performance Max

Audience signals in Performance Max tell Google’s AI who your best customers look like, accelerating the learning phase and improving conversion quality. Unlike standard campaign targeting, signals are suggestions rather than restrictions — Google uses them as a starting point and then expands to find similar converting users across all Google properties. This guide covers how to build effective signal strategies that drive better results from your PMax campaigns.

Key Takeaways

• Customer match lists are the strongest signal type — upload your best customer data first
• Signals guide Google’s AI but do not restrict it; the AI will expand beyond your defined audiences
• Layer multiple signal types for each asset group to give the algorithm more starting points
• Custom segments based on search behavior are the closest thing to keyword targeting in PMax
• Review audience insights monthly to see which signals are driving conversions and which are being ignored

Introduction

Performance Max campaigns serve ads across Search, Display, YouTube, Gmail, Discover, and Maps. Without audience signals, Google’s AI starts with no information about who your ideal customer is and must learn entirely through trial and error — spending your budget to discover what works.

Audience signals short-circuit this learning process. By telling Google “these are my existing customers” or “people searching for these terms are likely buyers,” you give the AI a head start. The result is faster ramp-up, better conversion quality during the learning phase, and a clearer picture of which audiences drive performance.

The quality of your audience signals directly correlates with PMax campaign efficiency. Accounts with strong first-party data signals consistently outperform those relying only on Google’s pre-built audience categories.

Step 1: Upload Customer Match Lists as Your Primary Signal

Customer match lists are the most powerful audience signal because they tell Google exactly who has already converted. The AI uses these real customer profiles to find similar users across Google’s network.

What to upload:

How to create and upload:

1. Export customer data from your CRM or e-commerce platform
2. Format as CSV with columns for email, phone (with country code), first name, last name, country, zip
3. In Google Ads, navigate to Tools > Audience manager > Customer lists
4. Click ”+” and upload your file
5. Google hashes the data and matches it against Google user accounts

List size and quality:

• Minimum: 1,000 records for matching to work reliably
• Recommended: 5,000+ records for strong signal quality
• Google typically matches 30-60% of uploaded records to Google accounts
• Higher match rates come from including multiple data points (email + phone + address)

Refresh schedule:

Update customer match lists at least monthly. Stale lists miss recent customers and include churned users, degrading signal quality over time. Automate the upload process if possible through CRM integrations or scheduled exports.

Privacy considerations:

Customer match requires that your data was collected with proper consent and in compliance with applicable privacy laws. Google’s customer match policy requires that data was collected first-party, with user consent, and in a context where users would expect advertising use.

Step 2: Layer Website Visitor Audiences

Website visitor audiences provide behavioral signals — these are people who have already shown interest by visiting your site, viewing specific products, or starting a conversion action.

Create segmented visitor audiences:

How to create website audiences:

1. In Google Ads, go to Tools > Audience manager > Audience lists
2. Click ”+” and select “Website visitors”
3. Define the audience based on URL rules, page category, or custom events
4. Set the membership duration (30-90 days depending on your sales cycle)
5. Wait for the audience to populate (requires the Google Tag on your site)

Using visitor audiences as PMax signals:

Add your highest-intent website audiences (cart abandoners, product viewers) as signals in your PMax asset groups. This tells Google to prioritize users who behave similarly to people who have already shown purchase intent.

Important distinction: Adding website visitors as signals does not make PMax a remarketing campaign. Google uses these audiences to find new users with similar profiles. To actually remarket to these visitors, you would need a standard Display remarketing campaign or ensure PMax’s audience expansion reaches them.

Step 3: Build Custom Segments for Intent-Based Targeting

Custom audiences (called “custom segments” in the PMax interface) let you define audiences based on search behavior, app usage, or website visits. They are the closest equivalent to keyword targeting in Performance Max.

Custom segment types:

Creating search-based custom segments:

1. In Google Ads, go to Tools > Audience manager > Custom segments
2. Click ”+” to create a new segment
3. Select “People who searched for any of these terms on Google”
4. Enter 10-15 high-intent search terms relevant to your asset group
5. Name the segment descriptively

Best practices for search terms in custom segments:

• Use commercial-intent terms (“buy,” “pricing,” “vs,” “alternative to”)
• Include your product category and close variations
• Include competitor brand + product terms
• Avoid overly broad terms that describe casual interest

Creating URL-based custom segments:

1. Same creation flow, but select “People who browse types of websites”
2. Enter competitor URLs, industry publication URLs, and review site URLs
3. Google identifies users who visit similar sites and content

Combine segment types:

The most effective custom segments layer search behavior with URL behavior. A user who both searches for “project management software pricing” and visits Asana’s website is a high-intent prospect.

Identifying the right search terms and competitor URLs for custom segments requires understanding which queries and competitor interactions correlate with your conversions. Lyra’s PMax Asset Diagnostics analyzes search term data and audience insights across your PMax campaigns to recommend custom segment terms based on actual conversion patterns.

Step 4: Add In-Market and Affinity Audiences

Google’s pre-built audience categories provide additional signal layers based on aggregated user behavior data.

In-market audiences:

These are users actively researching or comparing products in a specific category. Google determines this based on recent search behavior, ad clicks, and content consumption.

How to select in-market audiences:

1. In the PMax asset group signal editor, select “In-market segments”
2. Browse or search the category hierarchy
3. Select categories relevant to your product

Common in-market categories by industry:

Affinity audiences:

Affinity audiences represent long-term interests rather than active purchase intent. They are weaker signals for direct response but useful for expanding reach.

Layer signals strategically: Start with your strongest signals (customer match, website visitors) and add in-market as supplementary context. Avoid overloading with weak signals (affinity, demographics) that dilute the quality of your signal mix.

Step 5: Monitor Audience Insights and Refine

Setting up audience signals is not a one-time task. Google provides audience insights that show which segments are actually driving conversions versus which are being ignored by the algorithm.

Accessing audience insights:

1. Navigate to your PMax campaign
2. Click on “Insights” in the left navigation
3. Review the “Audience segments” section
4. Look for the breakdown of which audience types drove conversions

What to look for:

Refinement process (monthly):

1. Export audience insights data
2. Compare conversion rates across signal types
3. Remove signals that consistently underperform
4. Add new signals based on:
   • Audience segments you see in insights that you did not explicitly signal
   • New customer data (updated match lists)
   • Competitive intelligence (new competitor URLs for custom segments)
5. Test one signal change at a time to isolate impact

Scaling audience signal management:

For agencies running PMax campaigns across multiple accounts, maintaining signal quality — especially customer match list freshness and custom segment relevance — becomes a significant operational task. Lyra’s PMax Asset Diagnostics provides cross-account audience signal health monitoring, flagging stale customer lists, underperforming signals, and missing signal types.

Practical Example

A D2C fitness equipment brand configures audience signals for three PMax asset groups:

Results after 6 weeks:

Actions taken:

1. Expanded Home Gym customer match list to include single-purchase customers (not just repeat buyers) to broaden the lookalike pool
2. Added more search terms to Yoga custom segment based on converting queries from search terms report
3. Simplified Accessories signals — removed in-market audience since expansion was outperforming it anyway
4. Created a new custom segment using competitor URLs discovered in the audience insights

Common Mistakes

• Not providing any audience signals — PMax without signals works, but it spends more during the learning phase and may optimize toward lower-quality audiences. Always provide signals, even basic ones.
• Using only Google’s pre-built audiences — In-market and affinity audiences are generic. First-party data (customer match, website visitors) is always a stronger signal because it reflects your actual customer base, not Google’s category estimates.
• Setting signals once and never updating — Customer match lists go stale. Custom segments lose relevance. Market dynamics change. Monthly reviews keep signals effective.
• Conflicting signals in the same asset group — Adding “luxury shoppers” and “bargain hunters” as signals in the same asset group confuses the algorithm about who to target. Each asset group should have a coherent audience thesis.
• Confusing signals with targeting — Signals are suggestions. If you expect PMax to show ads only to your signaled audiences, you will be surprised by the audience expansion. Use audience exclusions (separate from signals) to hard-block specific audiences.

Lyra’s PMax Asset Diagnostics monitors audience signal health across your PMax campaigns, tracking signal freshness, performance correlation, and coverage gaps to ensure your signals continue driving efficient conversions as campaigns mature.