Microsoft Adds AI Diagnostics to Its Ad Platform

Microsoft is building AI-powered diagnostics into its advertising platform, using Copilot to identify why campaign performance shifted and whether conversion tracking is configured correctly.

In a blog post, the company outlined the diagnostic tools along with expanded LinkedIn targeting, creative controls, and a new attribution model, build on the AI Max for Search and UCP feed updates Microsoft introduced this week.

Copilot now Explains Performance Changes and Tracking Issues

Copilot inside the Microsoft Advertising Platform now includes two diagnostic tools. Performance Shift Root Cause Analysis, is designed to break down why campaign results change. According to Microsoft, it identifies whether fluctuations are linked to ad delivery, bidding strategy changes, or shifts in competition. Instead of only showing performance drops or increases, it surfaces possible causes behind those movements.

The second tool, Conversion Tracking Diagnostics, focuses on measurement setup. It checks whether Universal Event Tracking (UET) tags and conversion goals are configured correctly. The goal here is to detect setup issues early before they distort campaign performance data or reporting.

LinkedIn Profile Targeting Expands to Search Ads

Microsoft is also expanding LinkedIn Profile Targeting to Search ads in a pilot across the U.S. and Canada. The expansion adds Job Seniority Levels, covering 10 levels including CXO, VP, and Director, and Custom Company Lists that let advertisers target based on criteria like event attendance or account-based marketing prospects.

New Controls for AI-Generated Ads

The company is also introducing controls for advertisers using AI-generated ad copy in Performance Max and Responsive Search Ads. Term Exclusions let advertisers specify exact terms to omit from AI-generated ads, ensuring brand-sensitive or legally restricted language never appears.

Messaging Constraints go broader, allowing advertisers to define brand, legal, and product guidelines that AI-generated assets must follow. These features are being piloted globally later this month, with China excluded from the rollout.

Data-Driven Attribution Replaces Last-Click

Microsoft is rolling out Data-Driven Attribution, which assigns credit across the full customer journey rather than relying on last-click. Google made Data-Driven Attribution its default model after removing four legacy attribution models in October 2023. Microsoft’s rollout aligns its measurement capabilities with what Google has already standardized.