There are two distinct problems in commercial real estate workflow: finding deals and analyzing them.

The search layer is about discovery. Which properties are available? What are the asking terms? What’s the market look like in a given submarket? This is a structured data problem — you need listings, comp databases, market statistics. The tools that solve it are essentially sophisticated search interfaces.

The due diligence layer is about analysis. You have a specific deal in front of you. You need to understand it deeply: financial projections, lease structure, physical condition, title issues, environmental exposure, market positioning. This is a document analysis problem — you need to read, extract, validate, and synthesize unstructured documents.

These are different problems with different tools. A tool that helps you find a deal on LoopNet doesn’t help you analyze the deal once you’re in the data room. A tool that helps you analyze the offering memorandum doesn’t tell you what else is available in the market.

The MCP ecosystem for real estate is building out the search layer first. That makes sense — it’s the simpler problem. Structured listings data is easier to expose through a standard protocol than unstructured document analysis.

But the professionals doing serious deal work — the analysts, the associates, the boutique shops running acquisitions — don’t have a search problem. They get deal flow through relationships. They have a due diligence problem. They have too many documents, too little time, and too much at stake to miss something buried in an environmental report.

The search layer and the DD layer look adjacent from the outside. From inside the workflow, they’re completely separate.

Building for the wrong layer is a common mistake. Know which problem you’re actually solving. +++