Baldwin Realty

The Challenge

Baldwin Realty's prospecting process centered on identifying off-market commercial properties along specific high-traffic corridors — properties where ownership data, tenant information, and contact details all had to be researched and compiled by hand. For every corridor they wanted to work, a broker had to manually pull parcel data, research owners, determine who was occupying each property, find valid contact information, and initiate outreach individually.

The work was repetitive and time-consuming. When brokers were focused on active deals, corridors went unworked and opportunities were missed. There was no way to run the process at scale without adding headcount, and the manual approach made it impossible to stay consistent across multiple corridors simultaneously.

The Solution

We built an end-to-end prospecting pipeline that handles the entire process from corridor selection through outreach — without manual input at any stage. The broker draws a target corridor, and the system takes it from there.

The pipeline starts by pulling all parcels along both sides of the corridor from the Regrid parcel database. Each parcel then runs through a seven-step filtration process that removes properties that don't meet investment criteria — including a maintained skip list of national tenants that Baldwin Realty doesn't want to pursue. Only qualifying parcels make it through.

For each qualifying parcel, the system identifies the current tenant using a two-stage process against the Google Places API. Stage one performs a text search for businesses at the address, then runs spatial validation using the parcel's polygon geometry — matching business coordinates within a 50-meter radius of the parcel centroid. If stage one finds no spatially validated match, stage two expands the search radius and tries again. Properties that are vacant or undeveloped are flagged as “Unconfirmed” rather than returning a bad result. This process achieves 80–85% tenant identification accuracy.

Once filtration and tenant identification are complete, the system generates a CREXI-formatted CSV file and submits it to CREXI's data enrichment API. CREXI returns all available ownership and contact data for each address — owner name or entity, phone number, and up to three email addresses per property. The bot merges this enriched data into a master corridor tracker in Google Sheets, which it has direct write access to, and exports a KML file the broker can open in Google Earth to spot-check results visually.

From there, the system hands off to an Instantly AI email campaign. Outreach goes to all available email addresses on record — up to three per property — with an initial email on day zero and a shorter follow-up on day seven. If an email bounces, the system goes back into the tracker and marks the address as invalid. Every property in the tracker carries a status: email sent, waiting for response, or response received. When a property owner replies, the email auto-forwards to the broker's work inbox with the property address in the subject line.