Team evaluations
Powerful, and easy to misuse.
From the Agency Pulse, the AI evaluates an agent the way a real estate coach would, or the whole brokerage the way an executive consultant would. It is the most powerful feature in the module and also the one with the most ways to be misused.
The AI does not compute the metrics
They arrive pre-aggregated from the platform. The model scores by dimension anchored in those figures and writes the analysis, and it is forbidden from inventing numbers or writing variable names. That is why a report says «90% conversion, in line with the brokerage average» and not a string of symbols.
There is also a per-lead report
Separate from evaluating people, there is a report on the WORK done on one specific file, built from its timeline: what was done, when, and what was not. It is for reviewing a case you lost without arguing from memory, which is how those conversations get lost.
How to read it without doing damage
- The window matters: thirty days of someone with two closings is an anecdote, not a trend. The dialog warns you when you generate it.
- Share it only when you are going to sit down and read it with the person.
- Read it in full yourself first. If there is anything you would not sign your name to, do not share it: the report does not represent you, you do.
An «areas to improve» report sent with no conversation does the exact opposite of what you wanted. The tool cannot prevent that. You can.