> One weird result: 11% of user accounts have an @gmail.com address but were signed up without OAuth
Not weird at all. It's the number of users who don't like reporting to Google every time they want to use your product. If I were a user of your product, I might have been one of those 11%. I never log in with a third-party identity if there is an option to create a username & password, and if there isn't such an option, I usually just walk away.
Back to the main thesis of the article. Whether or not your business benefits from social logins probably depends on the nature of your business. Are you in the social gaming scene? You absolutely need Facebook integration. Do you sell an app that caters to mobile developers? Most mobile developers need to have an account with each of the major mobile platform vendors anyway, so it's reasonable to expect that people will use their Google accounts.
But the article you linked to at the beginning was written by Aaron at Mailchimp, and Mailchimp's main line of business has no obvious connection with any of the social login providers other than that they often send email to them. Mailchimp's most profitable clients are corporations, who spend serious money with their corporate account. They might even have specific employees who are responsible for the company's login credentials. It's not hard to see why those clients might not be attracted to social logins.
tl;dr: Generalization is difficult to do properly.
Completely agree with the reason why a user would use a gmail account without OAuth, I'm just surprised that 11% of users actually did it since it requires significantly more work to do so.
I also completely agree with the benefit from social logics depends on the nature of your business. For a B2B play, having users sign up through Google OAuth makes sense, especially for organizations that are using Google Apps to manage e-mail and identity already. Presumably this will also allow an IT administrator to manage the account (though I haven't dived deep enough to ascertain that)
Basically, choose your third-party logins properly to maximize growth through partner channels. If your app isn't inherently social, there still are great integrations out there that make sense.
1. Click on the red button
2. Log in to Google if you're not already logged in
3. Authorize your website
4. Done
Just creating an account:
1. Click on the gray button
2. Let your browser fill in your name and email
3. Let LastPass generate a random password
4. Done
It's not a particularly complicated workflow if you're used to it. This is a challenge for anyone who is trying to become a universal identity provider: they need to make it significantly easier than the decades-old alternative.
Not weird at all. It's the number of users who don't like reporting to Google every time they want to use your product. If I were a user of your product, I might have been one of those 11%. I never log in with a third-party identity if there is an option to create a username & password, and if there isn't such an option, I usually just walk away.
Back to the main thesis of the article. Whether or not your business benefits from social logins probably depends on the nature of your business. Are you in the social gaming scene? You absolutely need Facebook integration. Do you sell an app that caters to mobile developers? Most mobile developers need to have an account with each of the major mobile platform vendors anyway, so it's reasonable to expect that people will use their Google accounts.
But the article you linked to at the beginning was written by Aaron at Mailchimp, and Mailchimp's main line of business has no obvious connection with any of the social login providers other than that they often send email to them. Mailchimp's most profitable clients are corporations, who spend serious money with their corporate account. They might even have specific employees who are responsible for the company's login credentials. It's not hard to see why those clients might not be attracted to social logins.
tl;dr: Generalization is difficult to do properly.