Usability testing is an amazing way to keep your ego in check. Just when you think you’ve designed the most amazing user experience, be sure to watch some real humans try to use it (queue the fresh ego bruises). I’ve found over the years that the problems identified in user testing nearly always involve terminology/copy.
I was recently running a user testing session in which participants were viewing articles that had been recommended to them. They were using a breast cancer information site, and all of the study participants had been diagnosed with breast cancer.
Knowing that some users would want to save articles that were particularly helpful, and delete articles that were off the mark, the software provided both options to users. The problem I observed was with the terminology used for taking these two actions. One button said “Save,” and the other said “Archive.” In observing about 10 users, the vast majority of them weren’t sure what the difference was.
Study participants said things like:
“I guess ‘Save’ is for saving it, and ‘Archive’ is for when you really want to save it for a long time because it’s so valuable.”
“’Archive’ must be for the really good ones, whereas ‘Save’ is just temporary.”
This confusion initially struck me as strange given the ubiquity of “Archive” as a synonym for “Delete,” but I was reminded (again) that I am not the user! I use Gmail every day (where Google has done much over the years to confuse users about how to simply delete messages versus archiving them). Let’s face it… a healthy portion of email is garbage, and it makes no sense, ever, to “archive” garbage.
Where do we put our most precious national treasures such as the Declaration of Independence and the Constitution? In the National Archives, of course! This is not at all the same thing as a dumpster… and it’s no wonder that users get confused when a piece of software conflates the two.
Archiving typically is a “soft delete.” The items are not actually being deleted… they are being marked as deleted, and retained indefinitely. Given how cheap Internet-based storage has become, this is pretty standard practice in web software. The reasoning goes something like this: If, down the road, some horrible occurrence transpires that makes us wish we could bring back all those deleted items… thank God we only marked them as deleted instead of actually erasing all trace of them! In other cases there are dependencies in the data that make a “hard delete” problematic. For example, if a system user is deleted… references to that user by other data objects in the application (e.g. “created by username”) may cause problems.
How do we decide between a soft delete (archive) and a hard delete?
A key criterion to evaluate when choosing between a soft and hard delete is whether all instances of the data object have inherent value. In the breast cancer example, high quality articles about my health condition pass this test. I may delete an article that I don’t think is relevant to me, only to realize that it is relevant after a follow-up appointment with my doctor (we see this very thing happen all the time in user research for this website). It improves my user experience if I can go into my “deleted articles” and recover it.
Email, of course, is a counter-example. Here, a true delete is desired because so much email is junk. And it’s important to note that there’s a difference between “junk mail” in the sense of spam, versus email messages that I never want to see again (e.g. idle banter by co-workers about who’s ordering lunch). In the latter case I do NOT want Gmail to think it should start filtering out the senders… which might negatively impact office productivity. While many email messages have value, it’s not the case that ALL of them have value.
Is “Archive” the right word for a soft delete?
An application’s terminology and copy are critical to usability. It’s frightening how quickly poor terminology can lead users astray. On the flip side, I’ve watched seemingly hopeless user experiences be snatched from the jaws of defeat with simple copy changes.
If your application has a real use-case for a soft delete, make this clear to the user with the right word.
In all the user testing I’ve done with people outside of the software industry (i.e. “normal humans”), I’ve rarely encountered a person who equates “archiving” with “throwing away.” On the contrary, “archive” means to store something that’s valuable in a safe place.
While your own user research will inform terminology, words like “inactivate” or “hide” might be more in line with user expectations than “archive.”
Have a terminology suggestion? Think I’m full of it? Leave a comment!