14

In my experience, having a "logout" button/link of some description in the upper right hand corner is (or was) pretty standard. Is this documented/recommended anywhere?

Sites that require javascript have been drifting toward hiding the link in a menu (albeit still a menu in the upper right corner) -- e.g., Stack Overflow, Facebook and Gmail. I've personally found this... violates the "Don't make me think" principle, to put it nicely.

Coincidentally, I just experimented: on SO, you can log in without javascript, but I can't find a way to log out!

So: are there any good documents to reference on this subject?

1
  • 1
    I would also have no doubts that logout should be placed in the upper right corner, it's a really common pattern. I see sites hiding the link in some menu that also have some sort of user profile which they would allow the user to maintain there. Sites that have user profiles are likely used more frequently, thus the user will learn quickly where to find the logout. Anyway, I would agree that this is not the ideal way. Commented Mar 22, 2012 at 23:04

6 Answers 6

7

Amazon is another example. I didn't notice this trend until you pointed it out.

Someone else asked the same question: Why would a web site hide the log out button?

Here's a short post on it that pretty much says:

By hiding the logout feature, you're more apt to simply close the browser or tab, but effectively remaining logged into a service. This allows Facebook to openly track your online whereabouts via advertising partnerships that all report back to Facebook.

Makes sense, as you pointed out that 'Login' is usually still very much visible. Sites like Facebook/Google have the luxury of violating principles (or perhaps, evolving them) because people will grumble but continue using anyway. I agree with user12999 that for the most part, you should probably stick with keeping it visible.

4

You asked for research citations...

Wichita State University's Software Usability Research Lab (SURL) published research into Expectations of Users’ Mental Models for E-Commerce Web Layouts and concluded:

Comparisons of the responses from users from four geographical areas worldwide show that, in general, participants had similar expectations on the location of the web objects.

The author also studied whether length of experience had any impact.

Note though that they only tested for Back to Home, Advertisements, Internal Links, External Links, Shopping Cart, and Help. The consistency of those results suggest the same could hold true for Log Out / Log In functions.

Another caveat is that that study was done back in 2000, and current conventions for the actual location of specific web elements may have shifted. The broader conclusion of common expectation probably still holds.

Heidi Adkisson also did research into common locations of web elements (2002).

Note that this didn't examine the user's expectations of element location, but instead looked at where websites were actually placing the elements.

Heidi also wrote an article Examining the Role of De Facto Standards on the Web.

1

I know this is an older string but wanted to weigh in with my experience on this.

We launched a mobile payments website with the log out link originally appearing in the footer. We discovered, however, that during the payment process (and other areas of the experience) users were fat fingering the log out option which was immediately kicking them out (my dev team dropped the confirmation box prior to launch).

When I went back to revise this process, I ended up creating a slide out panel and included the log out option in there. It's listed along with all other primary navigation options but is out of site to avoid fat fingering. If they're on the desktop breakpoint of the site, I rolled it under the user profile dropdown menu. It has helped by slightly cutting down the visual noise of the primary experience and got rid of the fat fingering issue.

0

I don't really see a problem with having the log out feature in a drop down menu of sorts, as long as the menu holds additional features as well. This is a convenient way of grouping administrative user actions. However, having only one feature (log out button) in a dropdown would suggest that you're hiding that feature to the user, even if that's not the intention.

I think the key point here is the placement, which more often than not (as you mentioned) is in the upper right corner. I believe that this design pattern could have its origin from the fact that the close button for Windows's windows also is in the upper right corner. Which would for many users be the intuitive area to search for the feature to close/log out an application.

I do realize that Mac OS has the close button in the left corner but there's no arguing that Windows still is the dominating platform, which could very well be the reason behind the convention.

0

It only violates the "Don't Make me Think" principle if people are looking for how to logout, but cannot figure out how.

I'm not convinced your average Facebook user has any interest in ever logging out, so why waste valuable real estate with a function only being used by security-conscious tech nerds?

3
  • 1
    Internet cafes are becoming ubiquitous, or at least Internet kiosks. The two corner grocery stores near me both have two or three computers set up which they rent out, as do the three camper van outlets, and the hotel.
    – Erics
    Commented Mar 24, 2012 at 9:17
  • I work at a University, I understand the dangers of the public terminal, but that's another edge case. What % of users are coming in form those? Commented Mar 26, 2012 at 13:44
  • Backpackers. They're as common as cockroaches, and they just love Facebook.
    – Erics
    Commented Mar 26, 2012 at 15:07
0

I am not sure of the reason behind the debate on the logout button here. I mean after billions of users via Facebook / Google are now acquainted to use the logout in a certain way, this unorthodox way has itself become a standard.

Most importantly, the end customers using the web-applications are anyway going to be part of this huge community and they already know it.

Hence, by not having this feature (hiding the logout) now in our web-application we tend be outdated by many and some might even ask for it too.

1
  • On the other hand: my userbase is specifically web communities that predate facebook and gmail; in many cases, these communities don't use either property and may even be actively opposed to them. Some (most?) audiences have lots of facebook/gmail experience, but not all do. You're right: to go with current trends, mimicing facebook/gmail makes sense. But that doesn't work for every niche. Commented Oct 22, 2012 at 15:32

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.