The score itself

What it represents

The score is a number from 0 to 100. It is the model’s estimated conversion likelihood for that specific record, expressed on a 0-100 scale.

It is not a grade, and it is not a statement about how sure the model is of itself. A record at 84 is one the model places high in the ranking, based on how records with a similar profile behaved in your own history.

Bands sit on top of the number for readability: hot is 80 and above, warm is 50 to 79, cold is below 50.

How it’s calculated

A model trained on your historical records estimates a conversion likelihood between 0 and 1 for each record it sees. That value is multiplied by 100 and reduced to a whole number. Nothing else is layered on top · no manual weighting, no adjustment after the fact.

Every score can also carry its factors: the fields that moved this particular record up or down. Those contributions are computed in the model’s internal scale, and they do not add up to the score. A record at 82 is not “40 points of job title plus 42 points of industry”. The factors tell you direction and relative weight, and that is all they claim.

One honest limit on the level. The score is built to rank, and ranking is where it is strongest. Whether 70 corresponds closely to a 70% rate depends on your dataset and how well the model is calibrated on it, so treat the number as a position in a queue first and a rate second.

What it means for you

Worked example: two leads come back at 84 and 41 from the same model. Working the 84 first is well supported. Reading the gap as “twice as likely” is not.

Scores are comparable within one model. Across two different scores, or two models bound to different rules, the numbers sit on separate axes even though they share the 0-100 range. Compare a lead to other leads scored the same way, not to an account score that happens to read 60.