34
12 Comments

Social login — Do users prefer Google, Twitter, or Facebook? Here is what we’ve learned on Cruip.com

When we started Cruip we weren’t sure if we wanted to offer traditional email/password signup, or social logins to authenticate our users! In the end, we opted for just social logins (via Google, Twitter, and Facebook), and 30,000 users later, this is what we learned:

Google, Twitter, or Facebook, what do users prefer?

According to our stats, users prefer Google (70.97%), followed by Facebook (19.69%), and then Twitter (9.32%) to sign in to Cruip. Here are the numbers:

  • Total users: 33.999
  • Login via Google: 24.136 (70.99%)
  • Login via Facebook: 6.695 (19.69%)
  • Login via Twitter: 3.168 (9.32%)

Social-Login-Stats-1

These numbers can be different if you have a different product and target, but overall, what we discovered is that users tend to not login with their favorite social networks, so that’s why Twitter ended up being as the less popular option on Cruip— our target (startups, founders, and developers) love Twitter and use it on a daily basis.

So, are Google and Facebook the best options for our users?

Certainly for the users, but not for us. Why? Unlike Twitter, Facebook isn’t very popular among our target, so they choose it as the “safest” option to sign in to Cruip. The result? Approx. 1 out of 10 users who sign in using Facebook has an absent or invalid email — basically, they haven’t used the social for years.

Users without email (numbers):

  • Total: 897
  • Google: 0 (0%)
  • Twitter: 27 (0.85%)
  • Facebook: 830 (12.39%)

Social-Login-Stats-2

On-boarding a user without an active email is useless, so Facebook is safe for our users, but a problem for us!

In the end, Google is the best option for both of us

Regardless of whether your app is offering Facebook, Twitter, or both, your users are going to sign in using the social they care less, so 30,000 users later, and looking back we would have used Google as the unique authentication option on Cruip!

Last, but not least, Google provides 100% accurate data, in comparison with Facebook and Twitter.

Final notes

  • If your business is B2B a traditional email/password signup si absolutely mandatory over the social login options.

  • The use of social logins can make the registration (or subscription) process faster, but you also lose control over a third party, so I would consider all the pros and cons before enabling it to my web app.

  • One of the most common problems with social logins is that users/customers often forget which social login they’ve used to sign in! We’ve solved this issue using this little Trick!

I hope this post has given you some pointers if you’re considering to enable social logins to your app :) If you’re more questions or curiosity, please let me know by replying below!

  1. 3

    Just out of curiosity, I've pulled some numbers from tasklog.app

    • E-mail: 57.68%
    • Google: 28.66%
    • Facebook: 12.67%
    • Twitter: 0.98%
  2. 1

    I love this type of transparency - thank you.

    However......... I feel like you can't have a fair analysis of what's best without offering native (non-social) login.

    I would wager that if you offered native signup you would have more users signed up.

    1. 1

      Absolutely, Mick, in fact, it was more a comparison (in terms of performances) between these 3 social login options! I agree with the native sign up - we've never given it a try because we should change our entire onboarding logic!

  3. 1

    Thanks for sharing your stats.
    May I ask if you made some A/B testing by letting only Google login (this way, you'll get 100% of emails)
    Because maybe people that have selected facebook in the first choice, can login via google (if there is no other option)
    Just curious...

    1. 1

      You're welcome, Tim! It's a nice point, but unfortunately, we can't! If we leave just Google people who have previously signed in using Twitter and Facebook can't log in to Cruip!

  4. 1

    Interesting stats, thanks for sharing Davide!

    Especially the one with many users using Facebook login not having a valid email address. I guess you'll be removing Facebook?

    Would be interesting to test email too as in some cases (especially recently with all the data scandals) that may be the preferred option.

    1. 1

      You're welcome, Mark! Unfortunately, we can't. As explained above if we get rid of one of these options, people who have previously signed in using Facebook (for example) can't log in to Cruip! It's a problem because there are also paying customers there.
      In regards to the traditional email/password, I totally agree!

  5. 1

    For my webapp https://Watermark.ink (it's an online photo watermarking tool) maximum users prefer email login, next comes Facebook login. Very rarely users tried Twitter.

    And your trick to solve which social login previously used is great, I wish more sites use that.

    1. 1

      It's very interesting that you're onboarding the majority of your users via traditional email/password option! Is Facebook performing better than Google in your case?

      1. 1

        I think so, I will do a proper analysis and post the results.

  6. 1

    Great insights, thanks.

    Showing the last login is a great little touch. Minor, seems obvious I'm retrospect, but really is not. And solves a real problem. Kudos!

    Are you going to use Signin with Apple as well? That's an amazing+safe experience from a users point of view.

    For anyone who's working on adding login (incl. SSO), I would recommend strongly considering something like Auth0. Really streamlines the process.

    1. 1

      You're welcome, jki! We've never considered enabling Apple, but you made me curious to search for more information :)

Trending on Indie Hackers
I’m a developer who resisted doing marketing for years. Here’s what made me change. 31 comments Recently sold a bootstrapped B2B SaaS for more than $500k. AMA. 20 comments A Tool to Quickly Scaffold Custom SAAS Projects 17 comments We Built AppSumo’s Partner Experience Team from Scratch (Here’s How It Happened) 6 comments How to start a podcast in 2021? (Part 1: Idea & Format) 5 comments We bootstrapped our SaaS to $50k MRR with just me and my co-founder - no employees 4 comments