Login usability suggestion

(Ryan Feeley) #1


Ryan from Firefox Accounts UX here. I have heard of a few people accidentally clicking on the Log in with Firefox link, and wanted to suggest the following changes which I think will reduce this from happening as much in the future.


(Henrik Mitsch) #2

Hi @rfeeley, thanks for reaching out. Looping in @mbranson (IAM UX).


(Henrik Mitsch) #3

@rfeeley, it’s me again … some of us in the Open Innovation team are traveling over the next few days. It might take a bit until we get to respond.

Please accept my apologies for the delay.

Best regards,

(Megan Branson) #4

Thanks for the feedback @rfeeley. Can you provide some clarity as to the use case where people were accidentally clicking on Log in with Firefox? It’s not immediately clear why the proposed solution would reduce the accidental clicks; though I’m assuming it is the use of the word “Account”?

Some background, this is an ongoing back/forth conversation around language and one I believe will only be resolved with some A/B testing. :slight_smile: Looping in @hidde as this is a test we could look to execute sooner than later.

(Ryan Feeley) #5

Yes, exactly, adding the word “account” will make a difference. Most users are using Firefox, and may click it quickly without considering. Even I did it once!

(Megan Branson) #6

Thanks for the clarity @rfeeley.

We’re at a point of iteration with NLX where any further changes of this level need to be validated with testing and data. Otherwise we’re reactively rolling it back to a previous UX iteration that also had similar bug reports of the inverse nature.

We do have this logged as a suggestion and will look to test it against the current iteration! :slight_smile:

(Hidde de Vries) #7

I remember we did implement something not too far from this proposed solution before. We also tried ‘Log in with Firefox Accounts’ before, but that made the button stand out weirdly as its text was much longer than the others. I personally like the current iteration, but see how it could confuse some.

Agreeing with @mbranson that the best course of action here would be to run A/B tests to verify our intuitions with people.