On Requiring Facebook for Login

Last night, I had a great conversation on Twitter with Jeff Croft about the pros and cons of requiring a Facebook account for login. It’s a trend that seems to be on the rise and I, personally, don’t think it’s a good long term strategy.

It all started when I visited Home Elephant on a tip from GOOD magazine. It looks like an interesting service, but, as I have chosen not to create a Facebook account, is not something I can sign up for. And so it sparked this tweet:

AaronGustafson reminds you that not everyone has (or wants) a @facebook account. You’re limiting your reach by requiring one for sign up. /cc @HomeElephant

It prompted the following response from Jeff:

I’ll remind you that the limited reach may be a perfectly acceptable business decision, given other trade-offs.

To which I replied:

AaronGustafson agrees, @jcroft, a business may want to limit signups when starting up, but is 500 million potential users really a limit?

Jeff clarified his position (this was a series of tweets I’ve combined here so it’s easier to follow):

No, that’s not what I meant. What I meant was: Facebook offers you things you can’t get elsewhere. Those might be essential to your product. Or, it may not be worth the effort to build a version of your product that works without them. You said, “you’re limiting your reach.” I’m saying, “yeah, but I may be increasing my bottom line.” Sometimes doing what it takes to support standalone accounts means putting an excessive amount of resources into something.

I agreed with his position… to a point:

AaronGustafson sees your point, @jcroft. Still, it seems like there are services that offer options in addition to Facebook if you’re looking for shortcuts

To which he responded (combined from a short series of tweets):

What other service offers a social graph and auth? And if I used it, would you just complain that I require THAT service? If I need to accommodate people who don’t want to use Facebook (or whatever), I now have to build my own auth and graph which may be too costly. It’s a business decision. Maybe I’ll trade the extra users for the cost associated with them.

My reply (again, combined from two tweets):

AaronGustafson wouldn’t complain about a service that allows people who use different services to register/log in @jcroft. @janrain, for instance. AaronGustafson used @janrain in building the @StandardsSherpa site (in addition to supporting local system accounts). There’s a free version.

His response:

All I’m saying is, you have to balance business goals with technical and user goals. Multiple sign-ins is a user goal.

In a little cross-conversation, I replied

AaronGustafson agrees about finding a balance, @jcroft. But from a user’s perspective, being excluded from a service because you don’t is a turn off.

Jeff continued (again, from a series of tweets):

JanRain is great, but it’s only for auth, right? Doesn’t have graph, events, pages, groups, photos, etc. that FB has. If those things are essential to your app, you either use FB, or build them yourself (again, increasing cost). Really, I’m just making an observation that us UX people often forget that there are business goals, as well, and sometimes they conflict with the UX goals. Sometimes there are tradeoffs between optimal UX and cost.

And Jeff’s response to my earlier comment:

Yep, I agree. Just pointing out, we don’t know WHY that site requires Facebook. Maybe they totally agree that it’d be better if they didn’t, but doing so would have cost them millions. I dunno.

My response:

AaronGustafson agrees, @jcroft. There is a trade-off, but there are a lot of cons to @Facebook from a user perspective (many of which are privacy-related). They could have a valid reason, but it I think many companies take that route because they think “everyone” is on @Facebook

I then offered an anecdote (and a thank you):

AaronGustafson found that on @StandardsSherpa, with 6 login options + local, the split was 50-50 local vs oAuth. And that’s with a tech crowd. Aaron Gustafson enjoyed that discussion @jcroft. Thank you!

Jeff had a similar experience:

Yeah, I get the same thing on Lendle. About 50% sign up without FB or Twitter. But, it DID take me more time to allow that. For me, it was worth it. But if others come to another conclusion, that’s cool, too. And yeah, great discussion! :)

Anyway, I just thought it was worth preserving and sharing that conversation with all of you as login/auth and Facebook integration is a hot topic right now. Coincidentally, GOOD published an article yesterday about why one of their senior editors is not on Facebook.

Oh, and Home Elephant got back to me at the end of my conversation with Jeff:

@AaronGustafson Working our tails off right now for non-FB signups. Stay tuned…

Like it? Share it

Share on LinkedIn

Share on Google Plus

Comments

  1. It’s worth noting that Sam Gerstenzang also offered a great follow-up to my original tweet (which continued my thoughts perfectly):

    <blockquote cite=“http://twitter.com/ensuite/status/66643819805941760”>
    And even if one does have a Facebook account, not everyone wants their Facebook identity tied to other online identities</blockquote>

    Great point Sam!

    • Aaron Gustafson
    • | #
  2. This makes me realize I use the dreaded “series of tweets” too much. :)

    Thanks for transcribing the conversation!

  3. I await the day when OpenID achieves critical mass.

    • Andy
    • | #