Set Up Audience Signals on Performance Max

​

Overview

Audience signals in Performance Max help Google’s AI identify and reach high-value users by providing targeting guidance based on Innkeepr’s incremental prediction models. These signals don’t restrict targeting but guide optimization toward users most likely to convert incrementally, improving campaign efficiency across Google’s entire inventory.

​

Prerequisites

Technical Requirements:

• Active Google Ads account with Performance Max campaigns enabled
• Innkeepr integration connected to your Google Ads and GA4 account
• Sufficient conversion data for signal generation (minimum 30 days recommended)

Timeline:

• Signal setup: Immediate
• Campaign learning phase: 30-60 days minimum
• Performance stabilization: 60-90 days for reliable evaluation

Important: We recommend a minimum of 30-60 days for PMax campaigns to learn and adapt to new signals before evaluating performance. This allows sufficient time to collect data, optimize bid strategies, and stabilize across Google’s inventory types.

​

Setup Instructions

​

Campaign Structure

We recommend testing audience signals using a campaign split approach to ensure accurate performance measurement. Budget Allocation:

• Test campaign with audience signals: 20-30% of total budget
• Control campaign without signals: 70-80% of total budget

This split enables accurate performance measurement and direct comparison of key metrics like clicks, conversions, and ROAS at the campaign level.

​

One Asset Group per Signal

Create 2-3 asset groups targeting the same products, each with a different audience signal configuration. Different signal configurations allow Performance Max to optimize across various user segments and intent levels. Recommended Starting Configuration:

• Signal 1: 30-day lookback, Top 50%, No Conversion Lag
• Signal 2: 30-day lookback, Top 50%, 180-day Conversion Lag
• Signal 3: 90-day lookback, Top 30%, 180-day Conversion Lag

Note: If you are unsure which configuration to use, contact our team to determine the optimal signal configurations for your specific use case.

​

Apply Audience Signals

​

Configure Campaign Settings

​

Monitor & Optimize

​

Best Practices

​

Testing Recommendations

Campaign Duration:

• Minimum test period: 30-60 days for initial learning
• Recommended evaluation period: 60-90 days for stable performance comparison
• Allow Performance Max sufficient time to optimize across inventory types

Statistical Significance:

• Aim for at least 50-100 conversions per campaign before drawing conclusions
• Monitor confidence intervals when comparing test vs. control performance
• Consider seasonality and external factors that may affect results

​

Campaign vs. Asset Group Split

Google Ads does not provide click-based performance data at the asset group level through the API. Testing audience signals in separate campaigns (test vs. control) enables accurate performance measurement and direct comparison of key metrics like clicks, conversions, and ROAS at the campaign level. Why this matters:

• Asset group data is limited to impressions and conversions only
• Campaign-level data includes full funnel metrics (clicks, CTR, CPC, etc.)
• Cleaner attribution and easier performance analysis

​

Target ROAS Considerations

Set identical tROAS values across test and control campaigns. Google Ads favors campaigns with lower tROAS due to increased bidding flexibility, which can skew results and make it difficult to isolate the impact of audience signals. Best practices:

• Use your standard tROAS as the baseline for both campaigns
• Adjust tROAS only if both campaigns deviate from target by the same margin
• The actual tROAS value should align with your business objectives and historical performance

​

Inventory Allocation

Performance Max dynamically allocates spend across Google’s inventory (Shopping, Search, Display, Video, Discover) based on recent performance. This can cause short-term volatility, especially when introducing new signals. Expected behavior:

• Innkeepr-powered campaigns may shift to Display or Video inventory earlier than traditional campaigns
• Signals often meet conversion goals faster in Display/Video placements due to broader reach
• Search and Shopping inventory typically stabilizes after 30-45 days

If ROAS drops or spend shifts to less profitable placements:

• Increase target ROAS to push toward higher-quality inventory
• Temporarily reduce daily budget to slow optimization pace and allow recalibration
• Check the Performance Max placement report to understand spend distribution
• Allow 7-14 days after adjustment before evaluating impact

​

Common Pitfalls

Mismatched tROAS:

• Different tROAS values between test and control invalidate comparisons
• Always maintain parity in bidding strategies

Insufficient budget allocation:

• Test campaigns with < 20% budget may not exit learning phase
• Too little budget prevents meaningful statistical comparison

Ignoring inventory shifts:

• Performance Max will optimize toward the most efficient inventory
• Display/Video early performance is normal and often leads to strong overall ROAS
• Monitor placement reports rather than assuming Search is always best

​

Troubleshooting

​

ROAS Below Target After 30+ Days

Issue: Campaign consistently underperforms ROAS targets despite sufficient learning time Solutions:

• Increase target ROAS by 10-20% to push Google toward higher-quality inventory
• Review placement report: Check if spend is concentrated in lower-performing inventory
• Verify signal configuration: Confirm audience signals align with campaign objectives
• Check conversion tracking: Ensure all conversions are being captured correctly

​

Spend Concentrated in Display/Video

Issue: Majority of budget allocated to Display or Video inventory instead of Search/Shopping Solutions:

• Increase target ROAS to incentivize higher-intent placements
• Reduce daily budget temporarily to slow optimization and allow recalibration
• Wait 7-14 days after adjustment—Performance Max needs time to rebalance
• Consider that Display/Video may be driving efficient conversions: Review actual ROAS rather than assuming Search is better

​

Campaign Not Exiting Learning Phase

Issue: Campaign remains in learning phase beyond 30 days Solutions:

• Increase daily budget to accelerate conversion volume and data collection
• Reduce target ROAS if it’s restricting bid participation
• Ensure sufficient traffic: Verify that products/services have adequate search demand
• Consolidate campaigns: Too many campaigns with small budgets prevents learning

​

Low Conversion Volume

Issue: Test campaign generates significantly fewer conversions than control Solutions:

• Verify audience signals are applied correctly to asset groups
• Check signal size in Innkeepr: Ensure audiences have sufficient users
• Review signal configuration: Overly restrictive signals (Top 10%, 7-day lookback) may limit reach
• Wait 60+ days: Some signal configurations take longer to demonstrate value
• Consider broader signal parameters: Test Top 50% instead of Top 30%, or 60-day instead of 30-day lookback