TLDR
I think of the business cost of errors as being largely dependent on two things.
- the severity of these errors (ranging from mild inconvenience to large distrust and user-risk) and its emotional impact on users
- level of ease in which users can switch to a competitor’s platforms
In regard to recovery flows that protect these streams, I would base it on meeting the emotion that user’s feel, which depending on the application can range from mild frustration (Slack message not sending), to urgency (Uber lagging or failing payment), to panic and anger (unable to log into Bank Account).
More Specifically…
Slack errors don’t meaningfully decrease my usage because I’m bound to the platform through Stanford groups, so switching costs are high and the business cost of occasional issues is low. However, if errors are persistent then this could push people toward iMessage or other alternatives. Uber, on the other hand, has a very high cost of error because switching costs are near zero,in the sense that if a ride takes a long time to arrive, fails to load, or struggles to accept a payment, I would likely immediately move to Lyft. Those errors directly block revenue and destabilize the marketplace model, so Uber’s recovery flows must focus on speed and seamless rebooking to avoid losing rides. Banking apps sit in the middle in terms of switching costs (setting up with a new bank takes a bit of effort but otherwise users don’t feel super bound to their bank). But more importantly, banks sit at the extreme end of emotional severity such that even small errors can trigger fear, panic, and distrust. Because banking by nature must be as risk-averse as possible, the cost of error is extremely high, and recovery flows must be exceptionally clear, calm, and reassuring to maintain user trust. From a UI standpoint I can see considerations of these errors effect on people with especially Uber and BofA providing incredibly clear, calm, and reassuring language, color, and boxing in their recovery pages for reporting an issue.
