I can't tell you about the password, but we had a registration form which asked for the e-mail address. You could actually read the e-mail address (like normal) so people could easily see and correct their typos had they bothered.
The thing is, they didn't. The e-mail address is used for activation and we had huge numbers of unactivated accounts. I don't have the exact numbers any more, so this is from memory: when we added another e-mail field and checked if they were the same, the number of unactivated accounts halved.
Currently, the number of unactivated accounts is consistently between 10% and 20%, so that should give you an indication on how much it saved us.
I can only assume that for passwords the problem of typos is worse, since you can't proofread.
Comments
I can't tell you about the password, but we had a registration form which asked for the e-mail address. You could actually read the e-mail address (like normal) so people could easily see and correct their typos had they bothered.
The thing is, they didn't. The e-mail address is used for activation and we had huge numbers of unactivated accounts. I don't have the exact numbers any more, so this is from memory: when we added another e-mail field and checked if they were the same, the number of unactivated accounts halved.
Currently, the number of unactivated accounts is consistently between 10% and 20%, so that should give you an indication on how much it saved us.
I can only assume that for passwords the problem of typos is worse, since you can't proofread.