Allison Parrish is a user on You can follow them or interact with them if you have an account anywhere in the fediverse. If you don't, you can sign up here.
Allison Parrish @aparrish

wtf is up with this design pattern on web apps where there are huge text fields for username and password on the home page and so you type in your username and password only to realize that the button below it says "Create an account!" and then you have to hunt for the tiny "ˢⁱᵍⁿ ⁱⁿ" button in the upper right-hand corner in order to actually sign in to your existing account?

· Web · 97 · 167

@aparrish because existing users have already decided to come back, and a little discouragement is okay. new users won't sign up if they don't know how

@nightpool @aparrish what if it was the same form and the "sign up" and "log in" buttons were next to each other and the same size....

@shel @aparrish hmm. probably fine although I don't know if that's been studied as thoroughly as the normal link/form setup. gut feel says it's slightly worse because people are still gonna click on the wrong one by accident but not sure how much worse, could be negligable

@shel @aparrish @nightpool why have two buttons? existing e-mail signs you in, nonexistent proceeds to account creation.

Security-wise that would leak information of whether an email is already registered. Current design doesn’t pay attention to this either though.

@nightpool this is a reasonable explanation but it kinda makes me feel like I'm being held hostage by the app instead of like I'm a valued customer

@nightpool @aparrish it's transparent '''growth hacking''' bullshit, I hate it. treating acquiring users are more important than the needs of your current users is hands-down unethical

@martensitingale it's really, really not. it's just a consequence of when decisions get made for each type of user.

@martensitingale overall users as a combination of acquisition and retention

@nightpool yes, I think it's dangerous for services to try to maximize the size of their userbases, because it's both easy and effective to employ manipulative/coercive techniques in pursuit of that goal

@martensitingale I agree with you but a MINIMAL inconvenience for existing users compared to MAJOR drop-offs in new user sign ups is like, a very easy decision to make.

@martensitingale there's a huge leap from "it's easy to make costly long term trade-offs for short term growth" to "we should stop caring about the new user experience all together"

@nightpool the new-user experience is important, since every user will see it, but they only experience it once. I don't think there's at all a zero-sum exchange between ease for new and returning users (perhaps minimal tension, but having clearly labelled buttons for signup/login seems fine), but shoving a common workflow behind an extra page load (several seconds and a mental context switch) seems unnecessarily ungracious

@martensitingale @nightpool imagine: websites which provide you with the login screen, and then "sign up" and "sign in" buttons.

"sign up" takes you to a page where you can finish filling in the sign up form.

@martensitingale there are no zero-sum exchanges in engineering. I agree that there are solutions that provide minimal tension, but those solutions are less reliable and well studied then "signup form on landing page and log in link"

@nightpool @martensitingale Do you know of any studies that look at the impact of such design on the fraction of users who visit the site and then are still active after say 30 days? My guess is that the impact is pretty small since the users who wouldn't spend even a little bit of time looking for how to register are probably very likely to just drop off after registering.

@seanl not off the top of my head, but common wisdom is that acquisition and retention have independent rates (so getting 5% more acquisition doesn't affect retention rates, and does have the expected absolute-numbers increase)

@nightpool If you just treat them as raw rates and ignore how you're achieving changes in the rates, then sure. But let's say 10% of users who visit your site register and 50% of those are still using the app after 30 days, so the funnel to 30 days is 5%. If you were to make registration automatic by just sticking a cookie on their machine so they didn't need to do anything at all to have an account, do you really think that would result in a 1000% increase in people still active after 30 days?

@nightpool @martensitingale i understand your underlying point but there has to be a way to optimize for first-time users that isn't incredibly frustrating for returning users. the two use cases aren't zero sum. also consider that i (and others like me) are less likely to even sign up for a service that obscures the existing-user sign in

@nightpool None of that addresses the fact that it's specifically irritating in an insulting way to existing users.

@deutrino if you choose to think your own pride is more important then having welcoming communities then I don't know what to say to you.

@nightpool Read the thread. Nobody actually LIKES this design pattern. It's near-universally off-putting. Metrics are not god.

@martensitingale @nightpool @aparrish if you use multiple devices / clear your cookies / otherwise don't have a login session, you are worthless

@aparrish well I think there's a difference between "arriving on a site for the first time" and "going back to a site". retention is important and difficult to optimize for and deserves a lot of effort, but once you've literally typed into your web browser then like, you've already MADE the decision to come back.

as the saying goes, "acquisition happens on-site. retention happens off-site"

@nightpool @aparrish yeah, i....... also generally find this pretty irritating. ( especially on sites where, like, actually finding the log in part is difficult / requires typing in a different url. although they probably would rather you just ~stay signed in~. )


but, like, not irritated enough to, uhh, dooooo anything about it. 😣

@aparrish @nightpool same. but then most users use apps/auto-login, so they are usually never presented with that decision, i guess?

(answer: because the only kpi reliably communicated to ux and devs is the number of new signups and they don't actually care about providing the service of the application as much as they care about hockey puck user count graphs and gathering your data for resale and sending you e-mail newsletters)

@aparrish This pattern, and the one where you show the username and password fields one at a time with a superfluous click in between are driving me crazy.

@darth_mall @aparrish this is usually/sometimes because of enterprise/single-sign-on integration

as in if you type your organization's email address and it sees the domain part matches an SSO integration, it redirects you to the SSO page

a click is a bit much, I agree. At least some pages still just let you tab

@aparrish At least _sometimes_ they're smart enough to just log you in if you just put in your username and password and click "sign up" but the ones that aren't... AAAAAAAARRRRGGGGHHHH

My guess: they expect you use Facebook/Google to log in so you're "automatically" authenticated and don't need the link.

Or they assume most people never come back.

The former may lead to latter.

@aparrish it clearly shows that user acquisition is higher valued than user retention.

@aparrish I have like five accounts on CamelCamelCamel because it lets you set up an anonymous watch, but then if you go to sign in it gives you a little warning that it will create a *new* account, and I only see that warning after LastPass autofills and I click it and WHY WOULD IT CREATE A NEW ACCOUNT WHY

@aparrish I was literally JUST about to post a gripe about this.

Also sometimes in mobile "sign in" isn't even ON the page, you have to open up the ☰ menu.

@Nentuaby @aparrish I wonder if this helps to artifically inflate the number of accounts? So the company can be like "look how many users I have!"

@hummingrain @aparrish I don't think it needs to even be as deliberately manipulative of the numbers of that; it's just what you get when you relentlessly optimize for the easiest factor to metricize...

@aparrish I don't recall where it was, but someone last year I ran into an even worse version. Clicking on the sign in link would literally give you a new user form, existing users had to then click on a second sign in link on that form to get the actual sign in form.

I like the way that Crunchyroll (as an example) is set up with a combined sign up/sign in form. Fill out the left side to sign up, fill out the right side to sign in.

@aparrish I do not know...

I think they are really trying to push you toward 3rd party OATH authentication using Facebook or Google. Although I don't know /why/.

@aparrish I will say, that if I ran a small site, I'd appreciate using third party auth, so I wouldn't have to be responsible for storing and protecting login info.

@shoutcacophony @aparrish come to me you fresh delicious consumer data point

@aparrish As a UX person, I think there’s probably an elegant way to utilize the giant text fields for both login cases, and let the system detect a new sign up vs a login, then walk them through the (quick) sign up process, if needed, before passing them on to whatever they were trying to do.
Really, until the user is manipulating data tied to their personal account, login can be deferred until user authentication is absolutely necessary.

@aparrish when the user acquisition team has all the power and it's clear no-one gives two shit's about the dipshits already using the site when there's squirrels to chase

@aparrish masto picked up a (small) part of this design pattern and it irks me tremendously

@aparrish thank you I'm glad I'm not the only person this annoys.

@aparrish The irony is that after paying 100k+ for that design, they’ll then burn through 100k+ for a multivariate test to show them than a big, red button above the fold saying “sign in while you still can!” converts better.

@aparrish This is an onboarding thing that some startup somewhere implemented first and now gets copied everywhere. It’s really annoying to existing users, but it’s designed to make sure the new users have the fewest barriers

Apparently, most people are reluctant to leave what we know and are familiar with, but quick to discard what we don’t know. So they get away with it for the most part

All they care about are numbers so they can entice investors

@aparrish Ugh, I hate that.

And, like, this probably works if you're running a service very specifically for tech nerds. The kind of person who Googles "facebook login form" every time they want to get in is more likely to create a hundred accounts than ever work this out. Anyone in between? Who knows. Probably lose a few.

@aparrish i think it’s the same strategy as how landlords are extremely attentive to your personal needs up until the moment when you sign the lease

I'm sorry there are too many replies and I couldn't read them all but here I am

I was impressed with a company who combined the form; the form would log you in or create an account if you didn't have one.

Except I noticed this by accidentally creating a second account. So there's that too.

@aparrish caring for data aggregation and not user retrntion?, I guess?

I wonder why, this is so unnerving.