Lesson 03 · 5 min 5 min

Client insights

What their browsing says and they do not.

When you share properties with a client, everything they do leaves a signal: what they opened, what they saved, how long they stayed and when they were last there. «Client insights» reads exactly that.

First the numbers, with no AI

A temperature: «Hot», «Warm» or «Cold», computed from their behaviour.
Their favorites, their views, the time they spent and their last activity.
«What they want»: the features that repeat across what they look at.
«Properties of interest», with direct access to each one.

All of that is there without generating anything. They are objective signals the platform computes, and that is why they do not change depending on the kind of day you are having.

Then the read

«Generate AI analysis» adds two blocks on top of that data: «Client read», which says what is going on, and «Recommendation», which says what to do now. It is stored, so reopening the lead keeps it there; regenerating is your decision, not an automatic spend.

When there is nothing to read

If the lead has not browsed yet, the screen says so plainly and proposes the only thing that applies: share properties with them to start reading their behaviour. It does not manufacture a read out of zero data.

That is the detail separating this from a horoscope. With no signals there is no narrative, and the screen would rather be empty than wrong.

Tip: Look at the insight BEFORE you call, not after. Knowing they saved three houses on the same street and went back twice to one of them changes how you open the conversation entirely — and it is exactly the kind of thing a client will never tell you over the phone.

Previous lesson Next lesson