Earlier quoted context omitted.
"Where?!" Everywhere. It's being phased out in many places, but as a rule of thumb, mostly everywhere still. "known to be very bad ... been shouting ..." Right, yeah, to put it in some perspective remember that you're talking second factor here. This is not your login, this is a secondary confirmation and you still need some serious motivation to bypass it. It's definitely doable, I work in security and I know what k…
I'd never seen or heard of it being used for payments before, which is why I asked - I'd heard of phone numbers being used as account names (effectively) in some payment systems, but being involved in the workflow of making payments is entirely novel to me. I'm aware it's not your login, but it feels the same as asking someone for publicly searchable information to "verify your identity" - an additional "security" st…
SMS works with every conceivable phone, even most landlines if need be, users don't have to install a separate authenticator app, which may require a Google/iCloud password (now where did I put that post-it note?), that takes up space that may be scarce on low-end phones and that may not even be compatible with very old phones, leaving affected people in a really unsatisfactory spot.
Then they need to set up codes for every login, figure out how to switch back and forth between apps and how to copy codes, which is not very discoverable at least in Google Authenticator – most people seem to memorize and type instead, cumbersome.
Hardware tokens are even worse, people misplace those a lot and unless you are a bank with a mature process for issuing these, setup is probably even more of a hassle.
All of this may be big deal if you (also) target less technical people and want them to use your product when they have the option not to.
With SMS, all the user needs is a phone number. Pretty much everyone is familiar with that, most will readily share it, too. iOS will even extract codes and show them on top of the keyboard, just wait a second or two and tap the code, done. It's about as painless and frictionless as it can reasonably be, with apparently relatively inconsequential security drawbacks – given it's supposedly trivial to fake SMS, there don't seem to be a lot of people doing so at scale. Maybe a breach like this one will finally change that? Remains to be seen.
For now I can totally see why one might stick with SMS as a second factor.