Your visitors are already telling you who’s ready to buy.
Numzy scores them in under 60 seconds, on rules you write and can read line by line. No black box — every point is on the record, including the one we’re keeping on you right now.
What comes out
Three answers, and the reason for each one.
Numzy watches the behavior, adds up the points, and puts every visitor in one of three buckets. The threshold is a number you set, not a verdict handed down.
- Buy now
- ≥ 35
- Slack alert to the rep who owns the account, with the behavior that triggered it.
- Nurture
- 15–34
- Into the retargeting audience and the nurture sequence. Not handed to sales yet.
- Not a fit
- < 15
- Out of the ad spend. Suppressed rather than sold to — that’s where the budget goes back.
How it runs
One snippet in, a scored visitor out.
You embed our tracking snippet in your site or app. As visitors move through it,
events land in our Kafka-backed ingest pipeline. The scoring engine runs on
the event stream and classifies each visitor in flight, then emits
prospect_classified to whatever acts on it.
We store the results. You can audit them months later, when someone asks why a lead was routed the way it was. The records are yours — export them whenever you want, to your CRM, S3, or your data warehouse.
The scoring model
The whole thing is one file you can read.
No model to retrain, no vendor to ask why a lead was scored the way it was. Rules are YAML. You edit them, review them in a pull request, and ship them like any other config.
Change a number and the next event is scored the new way. That’s the whole deployment story.
Opening 2026
The ledger on this page is the product, pointed at you.
Every point it gave you came from the rules above, and you could check each one against them. That’s the promise for your visitors too — not that a score is never stored, but that you can always see how it was reached, and take the records with you.