This is a huge problem inherent in large-scale platform-type businesses in growth stages. The scale required to be competitive in these arenas is such that if support costs scaled linearly they would strangle you. At least until you reach near-monolopy status and can raise prices. These platforms each hit a nadir in customer service before a few embarrassing failures that trigger apologies and more transparency.
It's also worth noting that these customer service failures are only newsworthy when the platform has ubiquity. That is, getting banned only matters if it locks you out of a significant portion of a market.
As others point out there are some industries that have consumer protections. For example, Visa basically cannot single out individuals to be ineligible for Visa cards. It's worth considering whether these protections should be extended to any company controlling large market segments.
Maybe machine learning can (ironically) bring some humanity back to customer service. And I don't mean infuriating, worthless chatbots. I mean instead of having single-criteria lists of "offenses" that trigger bans have machine learning models that can look at customers holistically in a way that would be very difficult to manually program. This approach could be taken now using "dumb" scoring and weighting algorithms, but it would require a lot of manual work tweaking and analyzing.
There’s no evidence of reversal from nadir, ie it is an ever deepening black hole of “screw the customers.” eg. Google and Apple have no customer service. Generally the approach is fake apology then onwards into the abyss.
Why would Visa care to ban anyone from their network? They get paid no matter what, and have zero liability. The banks issuing Visa cards can certainly ban people, and use credit reports to deny applicants. Banks and merchants are the ones eating the fraud/chargeback losses, and they'll drop you in a heartbeat if you're causing them to lose money.
It's also worth noting that these customer service failures are only newsworthy when the platform has ubiquity. That is, getting banned only matters if it locks you out of a significant portion of a market.
As others point out there are some industries that have consumer protections. For example, Visa basically cannot single out individuals to be ineligible for Visa cards. It's worth considering whether these protections should be extended to any company controlling large market segments.
Maybe machine learning can (ironically) bring some humanity back to customer service. And I don't mean infuriating, worthless chatbots. I mean instead of having single-criteria lists of "offenses" that trigger bans have machine learning models that can look at customers holistically in a way that would be very difficult to manually program. This approach could be taken now using "dumb" scoring and weighting algorithms, but it would require a lot of manual work tweaking and analyzing.