Debt Management

Collections runs on more operational data than almost any other function and turns almost none of it into evidence. Contact outcomes sit in the dialler, arrangements in the servicing system, payments with the provider, and complaints and vulnerability notes somewhere else again. So treatment effectiveness gets argued from agent anecdote, and cure and breakage are recalculated by hand each month from exports that never quite line up.

We join those systems into one timeline per customer — every contact, arrangement, payment and outcome in order — in your own BigQuery, and carry the vulnerability and forbearance flags through with it rather than dropping them at the reporting boundary. The pattern that usually surfaces first is a treatment pathway that looks strong on cure but quietly high on breakage a few months out, which a point-in-time cure number was never going to show. Measuring by cohort and pathway makes the difference between a strategy that works and one that only appears to.

The output is fair-value and vulnerable-customer evidence with the working shown, and treatment performance you can actually act on: which pathways to expand, which to retire, and where a cohort is being harmed rather than helped. Cost to collect, contact-to-arrangement and promise-kept rates all read from the same events, so an operational decision and its regulatory evidence are the same number rather than two reconciliations.