in UX, Разни

Of UX antipatterns #1: Over-notifying

Users need feedback. They need to know when an action has finished, successfully or not, they need to see progress, and they need to know when a process is estimated to finish. 

Feedback can and should include text, imagery, sound, and vibration (pick some, not all). Depending on the context and the message that needs to be conveyed, feedback can be more subtle. Or, if a quick and critical action is required, it can be abrupt and require the user’s full attention.

It’s essential for any developer to understand what is important, and what is just crying for attention. A negative trend I’ve been seeing a lot lately, is over-notifying the user of what is happening.

Unsurprisingly, I’ve an Android example here. Imagine I want to turn off mobile data. I slide down the notification bar and tap the little globe icon representing data access.

No, that doesn’t turn data off, it sends me to a useful screen telling me how much data I’ve used so far during the current billing period. Nifty. Then there’s a slider at top right. I tap it to disable data. It turns off, and there’s a lot of feedback.

I see the little toggle slider animate to its off-state. Then the data usage panel disappears. Then I can see the globe icon has turned grey – a long time colour-code of something being disabled. Finally, a notification slides in, informing me I’ve disabled mobile data usage. And I need to swipe it off since it will stay there until I enable my data plan again. 

Now, different users require a different level of information, so the developer can hardly know how much info is enough. But then, why guess? Why not make some feedback possible to disable? I mean easily. 

Maybe there is an option to disable the data-is-disabled notification as well as the how-much-you’ve-used screen. But if they’re buried down somewhere, I (the user), am not going to look them up. I’ll keep getting slightly annoyed every time, but then again, all it takes is a gentle swipe to get rid of it. Just every single time! 

Then, is it too much to ask for a small “don’t show me notification again” prompt? After choosing not to show again, a little piece of text can let me know how to turn it back on if I ever want to.

Some companies just don’t test their usage flows with their potential customers. Sadly enough, some use this sort of information overload in the hope of getting more engagement from their clients.

Recently, I signed up for a web service that checks your grammar as you type. I wanted to try it out. They have a free version of their service where you can check your writing, but only get information on the main grammar mistakes you made. The paid version was claimed to detect “advanced mistakes”, and sure enough, nearly every text you entered would have a few of them.

If the free edition of their product you could see the number of “advanced” errors, but not they were. And if you sign up, you get to see them. You get to see a lot of false positives.

It turned out on top of actual grammar mistakes, the service advises on things its authors regard as bad practices. This includes an “advanced” error every time you use passive voice. Or when a sentence ends with a preposition. It admits these are not errors, but it still counts them. You cannot disable the reporting of those “errors.” It’s both wrong and annoying.

But it is a mechanism to lure people into signing up and then into making them think it can do more than it actually does.  It’s not a bad service at all: it does make a lot of sound suggestions and can catch real errors, but I can’t help it not to feel tricked. I’m not going to renew my subscription.

In my previous post, I said every developer should care about user experience. Maybe you are not a decision maker. But you should always speak up when a feature is about to be implemented in a way you think would not be optimally useful. And if you are a mere user, you ought to report the things that annoy you, or that are simply not informative enough.

Don’t shy away. We all want useful software.