Hacker News Reader

Comment by blakeyrat | Hacker News Reader

The problem isn't that it's a joke. The problem is that Google didn't do any kind of UX thinking behind the change. (It doesn't matter whether the change is serious or a joke, you can't just ignore the user experience.)

The new button performed a destructive action (because it blocked the recipient from responding to the email thread) without any kind of confirmation. Additionally, it was placed where it was likely to be hit by accident. (Apparently where the "Send & Archive" button used to be, although I don't have that on my Gmail theme.)

In short, they didn't test the damned thing. Even a half-hour of user-testing would have shown how easily the button was pressed by accident, and how devastating an accidental press was.

And again: it doesn't matter if the change is a joke or not, you don't push up a change unless you've done some basic UX work on it.

View on HN

"Send & archive" is a feature in gmail, but by default it's turned off (check your profile). The extra "hilarious" thing about google's dumb prank is that they basically replaced that functionality with the micdrop button, so anyone who actually was using that feature couldn't.

In general I'd say it's a bad idea to play jokes with email. Most people use it for important work, and many use it as part of their livelihood (not just for professional correspondence as in finding jobs but as a component of their day to day work). Just... don't fuck with that, it's unprofessional to mess with people. Imagine a car maker doing an over the air firmware update of their cars to disable the signal blinkers because isn't it funny how so many people never signal while driving?

>>The problem isn't that it's a joke. The problem is that Google didn't do any kind of UX thinking behind the change.

Actually no, the problem is that the joke they came up with is the type that requires UX testing in the first place, i.e. it changes the behavior of a product in a very noticeable and non-reversible (by the user) way. As opposed to something harmless, like a joke announcement or something.

Several years ago Blizzard announced the "Archivist" class for their popular game, Diablo 3. They put a good amount of effort into the joke, with a short gameplay video and everything. But it was a joke, and the class didn't actually make it into the game, and everyone had fun.

Good UX is every action should be reversible. And have confirmations before any destructive behavior.

However, that completely goes against the point of this joke. It's not a prank if they prompted, "Are you sure you want to send a Minion gif and mute yourself from this conversation?"

How does a confirmation go against the point?

You don't ask for permission before you punk someone.

The sender was not the person being punk'd; it was the recipient. The person who would have seen that confirmation dialog would have been in on the joke...

I agree that it would make it less funny, but the takeaway from that conclusion is that they simply shouldn't do this prank at all. A joke isn't worth harming gmail's UX like this.

> Good UX is every action should be reversible. And have confirmations before any destructive behavior.

Nah, people just click through. GMail has a much better model: mask destructive / un-undoable behaviour behind a short timer, make the action undo-able until the timer is up.

Is that not what reversible means?

In some sense, yes. I guess it depends on how you want to define your words.

Strictly, the action you start when you click `Send' is reversible for a few seconds, before it actually launches the nukes. But I'd rather call it `delayed' when talking to a human about it.

Delaying is a way to make some non-reversible actions sort-of reversible.

Hell, the minion gif itself would probably be fine. "Oh, shit, sorry: Pressed the wrong button on a dumb April fools prank."

Muting is a much bigger problem.

As someone who was interviewed (and rejected) for a UX role at Google, this incident doesn't surprise me. They seem very focused on A/B tests and don't seem to value qualitative research very much, and when they do it's usually survey based.

Reply on news.ycombinator.com