|
|
|
|
|
by hirsin
2847 days ago
|
|
What you're describing is a "cost proof" - namely that the user has something we can verify that costs some amount of money and is unique. So when the service I work on asks for a phone number verification, it's not always to determine your ID - it's to cut down on spam from users unwilling/unable to set up tens or hundreds of phone numbers, which I imagine is the majority of spammers. Adding it to existing accounts, though, makes less sense to me. Retroactively checking that an active account can cost proof seems like the most intrusive way of doing this, particularly as part of OS login - at this point you have so many signals that you should already be able to detect the user is a spammer or not. |
|
If anything I think it's the opposite --- dedicated spammers have shown they can farm resources like accounts of various types, so phone numbers aren't out of their reach. It's the casual users who don't want to give away their phone numbers or setup a throwaway one which will be turned away.