Sanaf, the AI employee

AI lead qualification

Every inbound lead looks identical in a form submission, so the team works them in arrival order and spends the afternoon on the ones that were never going to buy. This job asks the qualifying questions your best salesperson would ask, scores what comes back, and puts the ready ones in front of a person first.

What it does

Concretely

  • Asks the qualifying questions as the lead arrives, not two days later.
  • Scores what comes back against the criteria you set.
  • Reads buying intent from what the lead actually said.
  • Routes the qualified ones to the right person by your rules.
  • Records which source the lead came from, so the question can finally be answered.

Arrival order is the worst order

Working leads first-in-first-out treats a serious buyer and a browser as equally urgent, and since the browser often fills in a form faster, it frequently treats the browser as more urgent. Nobody chose this. It is what happens when there is no step between the form and the call, and it quietly sets the ceiling on what a sales team can produce.

Speed and sorting are the same problem

Qualifying within minutes works for a reason that has nothing to do with the questions: the lead is still thinking about you. A qualification step that adds a day removes the benefit it was supposed to create. Doing it as the lead arrives is what makes the sorting worth having at all.

It sorts, it does not turn anyone away

A lead this job cannot read confidently still reaches a person, flagged as unclear rather than silently set aside. That rule exists because the cost of the two mistakes is not symmetrical: a badly sorted lead wastes an hour, and a discarded good lead is invisible forever. Scoring is a suggestion about order, not a decision about who deserves a call.

Questions

Frequently asked

What criteria does it score against?
Yours. The qualifying questions and the weighting come from how your team already decides which leads are worth an hour, written down rather than held in one person’s head. A generic scoring model would be sorting your leads by somebody else’s business.
Does an unqualified lead just get dropped?
No. Everything reaches a person; qualification changes the order and the context, not who gets through. A lead it could not read is passed on marked unclear, which is more useful to a salesperson than a confident score that happens to be wrong.
Can it tell us which lead sources are working?
Yes, because it records the source alongside the score and what happened next. That is usually the first time the question can be answered with anything other than an impression, and it tends to change where the marketing money goes.