Operations2 min read
Alternate MetaData in India: when to spend, and what to expect back
Tracing is bought as a magic answer to unreachable accounts and usually delivers a partial one. Used with a rule about when to trigger it, it is one of the better rupees a collections operation spends.
A meaningful share of any aged book is not refusing to pay. It is simply unreachable: the number is dead, the address is stale, the employer has changed. Every contact attempt against those accounts is spend with a guaranteed zero return, which is what makes tracing interesting. It is not a recovery tool. It is a way to stop wasting the recovery tools you already have.
Trigger it on the right signal
The trigger should be evidence that the contact details are wrong, not that the borrower is not paying. Those are different problems and only one of them is solved by a new phone number. This is where a precise disposition list pays for itself: invalid number and wrong person are traceable states, while ringing and no answer are not.
- Number invalid or disconnected on two attempts. Trace.
- Answered by someone who says the number has been reassigned. Trace, and mark the old number dead so nobody dials it again.
- Ringing with no answer across several days and time slots. Do not trace yet. Change the time of day first, then the channel.
- Field visit returns address not found. Trace the address, which is a different product from tracing a number.
What you should expect to get
A partial answer. Not every trace returns something, and not everything returned is current. Treat results as candidates to be verified by contact rather than as facts to be written into the master record. The operational discipline that matters is what happens to a traced number that turns out to be wrong: it should be marked, not silently retried next month.
The conduct line, which is not optional
Tracing sits close to a boundary. Locating a borrower is legitimate. Contacting their relatives, employer or neighbours about the debt is a different act, and treating a traced third-party number as an alternative line to the borrower is where operations get into trouble. Decide the policy centrally, write it into the system, and do not leave it to a caller under pressure at month end.
Measuring whether it is worth it
- Hit rate: traces that returned anything usable.
- Contact rate on traced details, which is the number that matters, since a returned number nobody answers has changed nothing.
- Recovery on traced accounts against the cohort that was not traced.
- Cost per re-established contact, which is the figure to compare against continuing to dial a dead number.
Run those four for a quarter and the question of how much to spend on tracing stops being a matter of opinion.