Realtor.com automates search insights in seconds with AI agents built on Microsoft Advertising MCP

Since its founding, millions of buyers have used Realtor.com to find their dream home. While it's one of the most trusted brands in the real estate market, the company knows that the competition is only one click away.

Realtor.com relies on competitive keyword analysis to understand where competitors are gaining visibility in search platforms like Bing. But getting those answers often means pulling data from multiple sources, manually reviewing and routing findings, and frequently switching tools.

As the company explored how AI agents could automate that work by identifying and acting on recommendations, it needed to ensure workflows could scale and that human experts could still approve changes on the platforms.

This meant making campaign insights easier to access across the organization, so the team adopted Microsoft Advertising Model Context Protocol (MCP).

Rather than limiting advertisers to a single AI platform, MCP makes campaign data available within the tools and workflows people already use every day, allowing them to ask questions, explore performance, and spot new opportunities.

Working with Microsoft Advertising, Realtor.com created a scalable workflow that combines AI-assisted analysis with human expertise.

Here’s how the team did it...

• Connected data: They built Claude agents on top of Microsoft Advertising MCP, making auction insights data easy to access and creating a workflow that can expand into new use cases.
• Smooth handoffs: The team routed findings and recommendations from the AI agent to search marketers for discussion, then passed them to another reviewer for validation before any campaign changes moved forward.
• Shared prompt library: Building a shared library of prompts and skills enabled the reuse of successful workflows and expertise across accounts and use cases.
• Human control: They kept people responsible for every decision while AI acted like another member of the team—handling research, generating recommendations, and preparing work for review.

Realtor.com transformed competitive keyword analysis from a slow, manual task into a fast, AI-powered workflow. Insights that once took days to uncover can now be surfaced in minutes, helping the team act with greater speed and confidence. With the ability to audit tens of thousands of keywords at once, the team gained a far more complete view of the market without adding complexity or time.

Crucially, the workflow can be run independently without the need for dedicated data science resources. This empowers marketing teams to access and apply advanced analysis directly, while maintaining the human oversight required for enterprise campaigns.

The result is a repeatable, scalable framework that Realtor.com can continue to build on as it explores new AI-driven use cases.