Monday, October 24, 2011

Registration Form Design with Facebook, Twitter, LinkedIn Authentication

I continually face the challenge of designing and building registration / sign-up pages on a wide variety of different web sites and mobile applications.  Back in January 2010, I wrote a post that's one of the most popular on this blog:

When to Use Facebook Connect – Twitter Oauth – Google Friend Connect for Authentication?

That post looked at when and why you would use Facebook, Twitter, LinkedIn, etc. as part of your registration and authentication mechanism.  While some technical aspects in the post quickly changed (e.g., Facebook changed its data caching policy), many of the issues remain the same.  It's definitely worth a read if you've not seen it before as background for this post.

Design Challenge

In this particular web site, we needed to get the user's email (and password).  We also wanted users to provide information about twitter and LinkedIn to help personalize the application.  At first it seems like this should be pretty easy to design - think again.

Here's a classic example that illustrates the issue - the sign-up page from Mahalo.


Mahalo offers you the choice of sign-up / registration via a host of social networks (powered by JanRain).  They tell you the other option is to provide your email and password.

At the heart of this is two main advantages of having social network sign-up / registration. 

  1. Allowing the user to click on the Sign-Up / Register via a social network such as Facebook, Twitter, etc. means faster, easier sign-up and increases the likelihood of sign-up.
  2. By having the user sign-in to twitter or Facebook, we can ask for permissions to tweet or post to their wall or other social actions on their behalf.  We also can get access to their social graph.

For some solutions, it's sufficient to just have authentication via the third party, but in my experience it's pretty rare to have that be the only mechanism used during a sign-up process for a startup. 

In my prior post, I point out two major downsides of relying only on registration via a third-party social network:

  1. Multiple Social Networks.  If there are multiple social networks offered, returning users may not be recognized if they click on a different social network.  In other words, the first time I sign-up with Facebook.  At a much later time or on a different computer (no cookies), I come back and try to sign-in with Twitter.  The system won't recognize me.
  2. Email Address.  The other problem is that only Facebook currently allows us to grab an email address.  If you want an email, you will need to either limit your sign-up to be Facebook only, or you will need to ask for an email address after their initial sign-up.  In other words, you are likely still going to go through the registration that's on the right side of the Mahalo sign-up page.

This is exactly what Mahalo ends up doing.  I signed into my twitter account and gave them permission.  But now Mahalo must come back and ask me to either establish my account (New User) or sign-in as an existing user.


As a user, this feels a little strange.  I first click on the register or sign-up button.  I am presented with a choice of using a social network to sign-up OR providing my email and password.  I choose a social network, but then I'm forced to put in an email and password anyhow.  Huh?

I've tried a few times to design around this and do something else, but it doesn't really work out in practice.  And you will certainly see this design pattern all over the place.  Here's XYDO's sign-up box:


It's behavior is a little different, but also a bit funny.  At first it seems like I can sign-up with Twitter or Facebook.  However, if you sign-up with twitter, it puts a check mark next to it and leaves you on this page still requiring you to enter your name and email address.  Even with Facebook, where you can get a name and email, it still needs to ask you for your user name and password.

StoreEnvy has a sign-up with Facebook option on it's first page.  It grabs your email.  But it still is left asking you for more.



Bottom Line

In our case, we were designing an application where we needed to get the email and password, but we also wanted to get users to authenticate with LinkedIn and Twitter in order to get information about them and their social networks.

At first we tried down the same path as Mahalo.  We provided a sign-up page that had an alternative of signing up with a social network (twitter or LinkedIn) OR signing up by providing an email and password.  However, we ran into the same problem as Mahalo where our second page was then asking for more social network information and the email and password.  The confusion factor was just too high.  So, we went back to a more basic design.

Up front, we have a well designed sign-up / registration form.  You can see a bunch of great information on designing these forms in the following posts:

The second step of the sign-up process asks the user to authenticate to their Twitter and LinkedIn accounts.  This is more along the lines of a multistep sign up using a progress tracker.


In our case, the steps are:

  • Step 1 - Get email and password
  • Step 2 - Get social network information / authentication
  • Step 3 - Confirm additional information

I personally think this is a more "honest" approach and makes it much more clear to users what is going on.

Thursday, October 6, 2011

Customer Validation - 33 Great Articles

I saw a great post from Tristan Kromer Pivoting on Investor Feedback a.k.a Beware of Mentors.  It was a top post on StartupRoar on Tuesday.  His basic point was:

If someone, including me, tells you something isn’t a great idea and there’s no market for it there are only two acceptable responses. Either:

"That’s interesting. I’m going to take that thought out into the field and validate it with my customers."


"You’re wrong. I’ve spoken to dozens of customers, I have a validated customer persona, built an MVP to test key behavioral hypotheses, and the data doesn’t back what you’re saying."

imageToday I had two conversations with early stage startups (see Free CTO Consulting).  In both cases, I was the mentor telling the person that the idea, as I understood it was "not great."  One was an enterprise software product.  The other was a consumer play with possible viral growth.  Neither felt like a slam dunk.  Would it take more work to sell the enterprise product than they could make on it?  Would the consumer one get traction and be viral?  In both cases, my gut said that as framed it needed work.

In both cases, the founder and I brainstormed on ways to shift the product in order to get closer to what I perceived as something that would be great.

Neither had done enough customer validation to be able to say the second line that Tristan suggested.  There certainly had not been enough validation of the concept as it stood.  I would guess that most people who are providing feedback on your idea have a pretty good filter for what you have done to validate the concept.  If you've not Done Your Homework, it will be obvious.

In both cases, the answer was that the founder would go to find other ideas, turn those into paper descriptions and validate it with customers.  Customer Validation 101.  And that's the first response that Tristan suggests.

Now some helpful articles on Customer Validation, Customer Development, Lean:

  1. Six Idea Validation Tests to Pass Before Startup
  2. Customer Validation Really Starts with In-Person Interviews
  3. A Day in the Life of Your Customer
  4. How to Know If Your Startup Idea is the Next Big Thing
  5. Some Startups Forget to Validate a Business Model
  6. The Challenging Pace of Lean Startup
  7. How To Prioritize Feature Development After Launching an MVP
  8. The MVP is a Process not a Product
  9. Digging Deeper into Lean Business Model Canvases
  10. Wasting Time Validating Assumptions?
  11. Startups Need to Make Leaps of Faith, But Not Blindly
  12. Lessons Learned: Validated learning about customers
  13. Entrepreneurship as a Science – The Business Model/Customer Development Stack
  14. You Can’t “Feature” Your Way to Success
  15. The Phantom Sales Forecast – Failing at Customer Validation
  16. Customer Development and Marketplaces
  17. No Business Plan Survives First Contact With A Customer – The 5.2 billion dollar mistake.
  18. Vision Synching in a Lean Startup
  19. The Fallacy of Customer Development
  20. Entrepreneurs, Lower Investors’ Risk by Validating your Start-up Company’s Business Proposition
  21. Lessons Learned: What is customer development?
  22. How I Test The Market Validity Of A Product
  23. Customer Development Gut Checks
  24. Build a Path to Customers from Day 1
  25. Top 3 Ways to Fail at Customer Development
  26. Creating Startup Success – Customer Development + Business Model Design
  27. Customer Development and Marketplaces
  28. Hubris, Passion and Customer Development
  29. Free Customer Development Help –
  30. The Fallacy of Customer Development
  31. Hubris Versus Humility: The $15 billion Difference
  32. Less is More, More or Less
  33. Yes, but who said they’d actually BUY the damn thing?