I think that social vs non social logins probably segments your users into 2 different types of people. I consider myself to be a very savy Internet user - I don't trust social login. I don't think any good can come from connecting my accounts like that and I know that some site is going to do something inappropriate on my behalf. As users become more savy (read: cynical) I think you'll find the develop the same view point.

Additionally we found on our site that social login increased conversions for signups but those people are generally of lower value to our freemium product in the long run because they're less likely to want to pay.

I second the "absolutely not" view.

In fact I "deleted" my Facebook existence and if you want to force me into a social login, any social login, you just lost me.

Your mileage may vary, of course.

a. This is a plug: these guys have a stake in this kind of "social" This is no proof, no comprehensive data but just pretty charts.

b. post hoc ergo propter hoc fallacy committed

c. I trust Mozilla more https://news.ycombinator.com/item?id=5408735

As a user, I can say that I do not like social login. Also, page views is meaningless as is number of mins spent on your site.

It seems like a false dichotomy to say that users either like or don't like social login - in my own thought process it completely depends on the site and the social login offerings available. I am often annoyed at sites which essentially block content if you don't sign in socially (f6s.com for example), but I am quite happy to avoid the hassle of creating a new username and password if 1) I don't really care about the site I'm visiting or 2) if I can sign in with a social identity which doesn't compromise what I am getting out of the site. I have several different twitter accounts for various projects and interests, so I might sometimes just use one of those that isn't my personal account to take a few steps out of a login process. The fact that the OP (gigya) allows dozens of social network login options (arguably too many ;) allows easy social login with secondary accounts.

This is dumb. Show me pageviews and I'll show you a lot more "non-social login" people. I bet there are just as many (if not more) people spending the same amount of time on site as those logged in via social accounts. You've just got millions of bounces and quick looks bringing down the average for non-social.

I think the real answer is that users like choices and freedom. You will gain the maximum amount of users if you allow them to use social logins if they so desire, or allow them to create a stand alone username/password for your site. By doing it this way you don't lose the users who rebel at social logins, but you still retain the users who prefer social login. The concept of "why not both?" isn't that hard to grasp. I understand that implementing/linking accounts from numerous social logins as well stand alone accounts is more difficult for the developer, but that's the price of maximum growth.

We're working on bringing more social experiences to my employer's website, but the 3rd party interactions will eventually be only for the social login. Commenting, reviews, etc are all going to be done in-house this year, moving away from Gigya's tools. Their moderation tools are... subpar. And we need to have the comments locally anyway since we want to include them in the page's HTML for SEO purposes. Social login providers like Gigya certainly have their place - who wants to keep up with the constant stream of changes to FB, twitter, etc login APIs - but so many of them want to be the primary data source. This data is going to be the core of the platform going forward, so it's super important to own it all IMHO.

The article is a little biased. Indeed, users who use social login spend more time on their website. It doesn't involve a causal link between time on website and social login: perhaps users are not the same, with different habits of navigation.

And users who use social login will perhaps spend more time if it's not provided.

Interesting, but I would be more interested in causal relationships (I.e. growth rate or engagement delta). I think it's quite possible that social merely segments users by their already existing commitment levels: anonymous users have no commitment, social users want to make their commitment painless.

People doesn't like to create new accounts to login. That´s all.

The "social login" defeats the idea of net neutrality. It limits your privacy and gives corporations more power over your accounts. Not really my cup of tea.

When we implemented social login into our app. It did boost our conversions. Retention is another story but at least one goal is met right after.

No they don't. I've run a ton of usability tests for clients walking through this very issue. Almost no one will even consider using it.

Not this user.

They work for us so far for users' engagement to register. 96% of registration done through social APIs.

Any idea why? Could it be a design "issue"?

That is massively out of proportion with everyone else. I suggest you look into what you broke on your normal registration process.

It really depends on the application. For a casual use app, the social logins are the smallest barrier to entry for the type of user they are targeted at.

Any "friction", like having to enter a username, email and password is three more steps than a social login requires, costing these casual users.

So, I can believe it is highly effective for specific applications.

That's a lovely theory and all, but literally nothing has a ratio of social:real registration anywhere approaching 24:1. Those casual use apps you are talking about get 70% social if they have really bad registration processes. Under 50% if they have decent registration.

Do not vote this up. The title submitted to HN is everything that is wrong here.

It's important to note though that the type of things that this 'plug-in' does is social already. When you're logging in to Pepsi SoundOff you already know you may want to share something on your Facebook feed, or you don't care about what brands can see your information on Facebook. Each service should consider what their target audience is and work towards using that method.

To be completely fair though, I never let anything login with my Facebook account, I just don't want other sites to have access to that data. And I would probably never use something like what this place provides, so people who use the net like me would never show up in their numbers.

>This is a solid counter post to Mozilla's post

In what way? It doesn't offer anything at all to counter that post, and doesn't in any way support the notion that users like social logins. It just shows that users who used social logins spent more time on the site.