Why does it take days to reverse a failed UPI transaction?*
There are more than a dozen moving parts owned by half a dozen companies in a typical UPI transaction. Data captured at one end of the transaction does not reach the other end in the same form. The different parties involved have vastly different mindsets towards uptime, CX, and TAT / SLA, which, in turn, means they make vastly different investments in tools and processes.
As ET PRIME once reported:
UPI still has “too many moving parts” and a “poor architecture” compared to credit card payments.
A system designed for mass adoption needs to minimise steps, failure points. UPI makes it easy for a merchant to accept payments digitally without an upfront cost of a POS machine. But it fares poorly when compared with a card transaction. This is a consequence of too many moving parts and a poor architecture of UPI.
More in Why Two Factor Authentication Is A “Conversion Killer” & “Blood Pressure Booster”.
Like any chain, UPI is only as strong as the weakest link. As a result, it’s not easy to troubleshoot a failed UPI transaction.
UPI is an overlay over the IMPS rail. I have heard bankers say it takes 2–3 three months to troubleshoot a failed IMPS transaction. If a UPI transaction is actually reversed in a matter of days, I’d say that’s a great accomplishment.