A free playbook by The Aigent Lab

Deal Sourcing & Market Research

Deal Sourcing & Market Research · Playbook 08

The AI research stack that won't invent numbers.

A general AI chat tries to shortcut serious market research with training data, and it gets things wrong in ways that look right. The fix is a dedicated Project per submarket, grounded in real published sources, with instructions that block speculation entirely.


Why this matters.

Underwriting a submarket properly means reading recent macro reports, submarket-level data, transaction comps, the supply pipeline, and developer track records. On a slow week that's a full day of work before any specific listing hits your inbox. The fix is to swap 'ask the AI' for 'ask the AI inside a project that only knows what you fed it.'
The AI is not the research. The sources are. The AI just reads them faster than you can.

One project. Five sources.

01

Create one project per submarket

Name it after the specific area, not a generic label, so the model never blends two markets together.

02

Upload the five sources

Macro consultancy reports, submarket-level data, six months of transaction and rental evidence, supply pipeline data, and developer track records.

03

Paste the grounding instructions

Tell the model to answer only from uploaded documents, cite the source and page for every number, flag thin sample sizes, and refuse to speculate.

04

Ask in plain English

Once the rules live in the project instructions, a one-line question returns a calculated, cited answer without you engineering anything clever.

The grounding prompt
You are my research analyst for this submarket. Answer only from the documents I've uploaded. Cite the source and page for every claim. Flag any figure resting on fewer than 10 data points. If a number isn't in the documents, say so and name which source would supply it. Do not invent or estimate numbers that aren't there.

What it still gets wrong.

i.

Source quality is the ceiling

The analyst is only as good as the reports you fed it. Refresh sources every quarter.

ii.

Public transaction data has a lag

Records typically surface 30 to 60 days after closing. The trailing month is usually incomplete.

iii.

Fees and charges drift

A two-year-old service charge or fee schedule can already be wrong enough to swing a yield calculation.

iv.

It cannot walk the area

A community can score well on paper and still have a real problem nobody wrote down. Visit the area. The project is the prep, not the decision.


Next Playbook

Build a bull and bear case on any deal under contract

Read Next ↗