Is OpenId easier or harder for users?

Question

I'm wondering if I should use OpenId for my website. My first exposure to OpenId was StackOverflow, and I found it confusing that they only had a login link, yet no register link. Now that I've learned about OpenId though I prefer it over the regular way of registration.

I have a feeling that only a small percentage of the internet users know how to login with a third party account provider, and most would prefer just to create an account. It makes sense for StackOverflow to use OpenId since the target audience is tech-savvy, however my website caters to the general public.

Does anyone have any statistics or first hand experience with using OpenId versus regular registration?

Answer 1

You should have it only as option, not requirement. People just don't understand this concept and often don't trust it.

Many stackexchange.com clients (hosted stackoverflow) have learned this hard way. There have been so many complaints that stackexchange.com developers had to implement traditional username/password authentication as an addition to existing OpenID method.

Answer 2

I think these days a lot of sites have facebook login and a lot more people know how to use facebook than openid. If I were you I'd go with facebook. e.g. dailymile.com

Answer 3

Heres a tip, don't implement login at all. If the users don't have to login there is no concern over what system to use. Plus, it's easier for the users to not login, then to login. Instead store all the data with one account. Then you can even watch as people overwrite other peoples hard work in real time. Although, if you want to have login. Make the passwords only 1 letter in length, then the user doesn't have to remember very much, and even more so, you can make it only require a username to login. Eg

if(getPostedValueByKeyAssociation('username') == $username) then {
    set <logged_in> to valueByAssociatedSymbol(NEUTRAL - FALSE)
}

Regards from Mr. Wong, The space janitor, USS Enterprise

Answer 4

Several members of the OpenID Foundation have done that sort of user experience testing. I don't know, however, which (if any) of them have published that research. It's certainly the sort of thing that the Foundation should make available, as they make some claims and it'd be nice to see the numbers behind them. Ask 'em?