The Stripy Strudel's Journal

Once there lived a software application that didn't aspire to the mass market and was quite happy with half a percent of permanent users. Those were also rather happy with the application; most of them had been using it since very old versions and developed their own approaches, habits and comfortable customizations. Many of them liked that the application was different from its competitors, and that some of its workflows are unique and unparalleled. But even if the competitors implemented all the same workflows and features, such a user would hardly make a switch because fine customization of an alternative application to one's long-fostered habits to achieve the same level of comfort is a lot of work and not always fully possible. Then, at some point, the vendor of the application is no more satisfied with the loyal but small user base and wonders: our product is better than the competition in almost all respects, and yet we have less than a percent users! Why? The vendor starts researching and analyzing, and it turns out that most users of the competing products think that this application is too hard to use and configure, and in general designed for a technical person rather than a regular user. Oops, says the vendor. We've never had it in mind to design for some kind of technical people! How come our application scares the users? We've been only working on functionality so far, it's time to work on our user interface. Our application must be user-friendly! Friendly to the regular user, that is, to the one who up to now has been using the competing products, being scared by ours. And because those who managed to attract the users obviously have user-friendly interface, we should learn from our competitors. An unusually long time passes between the releases, and finally a permanent user who has been loyal to the application since, say, version 2.0 for MS-DOS, and who is pretty sure that he's the damn regular user, installs a the new version, say, X (that's 10.0 to the new trend). Here is what he finds:

• Menus, toolbars and keyboard shortcuts are organized in the same way as the competitors have it, including their traditional but illogical peculiarities such as putting “Preferences” in the “Edit” menu.
• Features unused by 80% users have been moved away to “Advanced”, “Special settings” and “Extra tools”. Especially so for fine-tuning options. The problem is that for every such option, the 20% who use it are different, and in fact, most users need at least one of the relatively unpopular features.
• Features unused by 99% users have been dropped.
• Toolbar buttons surviving the purge have grown in size and got text labels, hoping to be noticed.
• The application started thinking for the user and offering various suggestions, tips and auto-configuration. In severe cases it makes the application seriously slow. It's especially relevant when the user knows perfectly what he means but can't finish typing because the pop-up suggestions get in the way or the application is busy computing them.
• Wizards with one or two items on each step have replaced dialog windows with all those items at once.
• There are new features, too, but they are quite strange, don't fit the general spirit and break the design principles of the application. They clearly look like the Professional User Interface Design has been finally applied.
• New features have sonorous names that tell nothing about what they do, such as EasySnap for aligning objects on a grid or QuickLink for uploading files into a mobile phone. Telling the users honestly what it is would scare them away, so everything should rather be quick and easy, no stress.
• New features have functional limitations without technical justification. For example, only up to ten user-defined folders: the research shows that the notorious 80% will never need more, while ten fits in the allocated space on the screen without scrolling.
• Error messages have become as informative as “Some error has occurred”. Of course, 80% of the users have never understood the technical details that the error messages used to contain, but when such a user follows the advise to “consult the system administrator” and calls a friend from the remaining 20%, that guy can only advise: “Try changing some option”.
• Features for integration with other products or web services aren't implemented as generic mechanisms that could be specialized for any favorite product of yours, but are rather made specifically for the most popular product in its category, which is used, according to statistics, by… well, you know. For example, instead of the ability to use a webmail service for sending messages you get the ability to use GMail for that.
• The application offers periodically to upgrade itself and sometimes even downloads new releases automatically and installs them, bypassing the system-wide upgrade mechanism (in those operating systems that have one). At least now there's one application for Linux easy enough to upgrade — usually everything is so hard there that it's strange that people even survive.

Of course, this is a degenerated, absurd case. All of this hardly ever happens at once, and the very process of making the interface user-friendly is often distributed over several releases of the application. I don't mean any single application. I remember this happening with different applications and to different degrees. I'm not going to provoke flame wars by naming any examples. I don't know whether this has ever helped vendors win some “regular users” from the competitors, but those who had long been loyal to the application definitely got reasons to try an alternative when the favorite application isn't what it used to be anymore. Every time, having found myself as one of such users, I felt offended by the vendor claiming the revamped interface to be user-friendly. If that's friendly to the user, than who am I? По-русски: Ворчание о дружественности