A brokerage team's deal operation mapped end to end, with explicit stages and named owners, and a real-time dashboard created inside the Google tools they already use.
For one residential brokerage team, a single deal crossed six tools between pitch and close, and the only thing connecting them was the coordinator's attention. Agents tracked deals in notebooks and verbal updates, so the record ran a step behind reality. The rules existed and the data existed. A person was carrying information between them by hand.
The first deliverable was an operations map. The deal lifecycle became explicit stages, each with a named condition for advancing and a named owner for every hop between tools. Writing it down surfaced the engagement's hard constraint: the brokerage platform the team is required to use has no API, so anything automated has to work around it.
The build honored one more constraint. No new tools. The team keeps the software it already pays for and knows, so everything lives inside Google, where they already work. From the map I built a real-time dashboard derived from the team's own spreadsheets, starting with quarterly pitch conversion. Once the states were on paper, the machine's share of the work was obvious.
2026 · Q2
Trend over time
01020304050
0%10%20%30%40%50%
2024 Q32024 Q42025 Q12025 Q22025 Q32025 Q42026 Q12026 Q2
PitchesSigned exclusivesConversion rate
Where deals went
Every loss in the period. Choosing one scopes the pitch detail.