miatzSign in

Scoring inbound leads from your own pipeline history

Train a classifier on the enquiries your pipeline already records as won or lost, then rank each new enquiry by its resemblance to the ones that closed, so the first call of the day goes somewhere defensible.

Effort
Weeks of work
Skill level
Some technical skill
Organisation size
Small business
Value
Time saved, Revenue

Tools named for this

  • An export of historic enquiry records carrying their outcome
  • A gradient-boosted or comparable classifier
  • A ranked queue surfaced where the calling actually happens

What to check before you ship it in India

  • A score is used to make a decision affecting the person scored. Section 8(3) of the DPDP Act 2023 requires personal data likely to be used for such a decision, or disclosed to another Data Fiduciary, to be complete, accurate and consistent — which turns a duplicated or stale contact record into a compliance failure and not merely an untidy one.
  • Labels come from the enquiries somebody chose to work, so the model learns the team's past preferences and then returns them as a prediction about the enquiry.
  • A ranked list is a calling list. Under TCCCPR 2018 a voice call whose primary purpose is to solicit business is a commercial communication, and regulation 32 bars a business not registered with an access provider from making one at all. Registration, template and preference scrubbing come before the first dial, not after the model is tuned.

Sources

Every claim on this page traces to one of these, on the date it was read.