CRM & pipeline

Real Estate Recruiting Pipeline

A verified pipeline of newly-licensed agents, built from state license records rather than scraped guesses. It was built in four moves — find the data, process it into something usable, organise it, then turn it into a product somebody can actually work — and the finished tool still shows where every single field came from.

Real estate recruiting State license records 139 verified rookies Cape Girardeau → Kennett corridor

See it work

The problem

Recruiting agents right after they get licensed is a timing game, and the honest data is scattered. State license records confirm who's licensed and when, but they don't say who's unhappy at their current office or worth a call this week. A lot of recruiting lists fill that gap by scraping review sites and guessing — which means part of the list is stale before anyone dials it.

Even the confirmed part has a trap in it. A license record lists an agent's sponsoring office under its registered legal name, not the sign on the door, so a plain text match against a public brokerage list quietly misses real pairs.

What it does

The standout feature isn't a list of leads — it's a tool that tells you where every one of those leads came from. It is a pipeline that publishes its own uncertainty instead of hiding it: license records as the primary, verified source; the reasoning behind brokerage DBAs versus legal entities, a real gotcha in this kind of matching; and, the honest one, a section titled "What is still not verified."

The five tabs

Rookies — 139 verified, Gatekeepers, Original 94 names, Brokerages, and Sources & method.

Rookies — 139 verified

Every newly-licensed agent in the corridor, confirmed against state license records. Columns: Agent, Sponsoring office, Licensed, Stage, Next action, Priority.

Original 94 names

The old review-scraped list, matched against license records and shown with a Match confidence column instead of being treated as fact.

Sources & method

The tab headed "Where every field came from" — license records as the primary, verified source; the corridor filter explained; brokerage DBAs versus legal entities; and what is still not verified.

How we build it

Every data product we build runs through the same four steps. This one is a clean example of all four, so here is the whole route from a public filing to something a recruiter opens on a Monday morning.

1 · Find the data

First we work out which source actually knows the answer, and go to that one rather than the convenient one. Here it is the state real-estate commission's public license records — a filed, public source that confirms who is licensed and when. That is deliberately not a directory scrape or a review-site pull, because those go stale and nobody can tell you when they did.

2 · Process it into something usable

Raw records are almost never usable as they arrive. The statewide roster is narrowed to one strip of geography — the Cape Girardeau–Kennett corridor — because that is the travel distance a recruiter can realistically work. Then the messy part: a license record files a brokerage under its registered legal entity, not the name on the sign out front, so a plain text match against a public brokerage list silently misses real pairs. Reconciling DBA to legal entity was done by hand, office by office. The old 94-name scraped list went through the same treatment and was kept with a match-confidence grade rather than being treated as fact.

3 · Organise it

Clean data still isn't a tool. Every agent becomes one record carrying agent, sponsoring office, licence date, stage, next action and priority, and the whole set is split across five tabs — Rookies, Gatekeepers, the Original 94, Brokerages, and Sources & method — so each screen answers one question, instead of one big table answering none of them.

4 · Turn it into something you can use

The last step is the one most data projects skip: making it usable. It ships as a pipeline with stages and next actions, so a recruiter works it rather than reads it — and it carries its own provenance. The Sources & method tab shows where each field came from, how the matching was done, and what is still not verified, so nobody has to take the list on trust.

The numbers

What we'd do next

Next would be widening the geography beyond this one corridor — step one run again against a bigger roster — and continuing to close the gaps the tool already admits to in "What is still not verified" rather than papering over them.