Five times tech companies ignored the obvious & correct design solution at their peril
Also known as Bad UX Roundup Returns #2
I have a huge file full of accumulated UX fails from a rogue’s gallery of companies and products. Usually for Bad UX Roundup, I’ll just grab a few that look like they’ll be fun to cover. But sometimes, I’ll notice a pattern in the ones I picked, and just like that I’ll have a theme for my article. And today’s episode has a very important theme that the UX world direly needs to hear.
Some bad UX is the product of cost-cutting leading to rushed features with zero design, or cheap farshore designers who think browsing Drivvvel is equivalent to an HCI degree. Other times, there is an obvious business motive for the company to inflict bad UX on the user. But then there are a disturbingly large number of examples of bad UX with no clear purpose or cause. The solution to the problem was so obvious that one would expect even a clock-puncher could come up with it, and yet what was pushed live is so head-scratchingly stupid that we’re left asking why.
I have come up with 5 such examples for you to, uh, enjoy?
Adobe’s horrible tab UI
I don’t put Adobe products on blast that much anymore. No, it’s not because their products are better. It’s because I stopped using them. I haven’t created wireframes in Illustrator in 3 years now that Figma is a thing. But I’m still not fully liberated from the Adobe ecosystem yet and that sucks because Adobe also sucks.
Today’s example is the mystifyingly bad window-sorting system in Illustrator, Photoshop, and probably also XD but that program blows so hard I don’t even want to open it to find out.
Let’s say I’ve got three windows open.
I want to move the first window over to the middle slot, so I click the title tab and drag it to the right.
Then I let go of the mouse button and now it’s sitting in the middle. Simple, right? You’re probably thinking these tabs behave just like the tabs on literally every browser on the market for the last 15 years. Well you’re wrong.
Let’s say I’ve got a bunch of tabs open, like below. I want to move the one on the right all the way over to the left. Sounds simple…
Uh oh, as I was dragging the mouse leftward, I moved it 3 pixels down in the process, causing the tab to detach from the tab bar.
Then, when I let go of the mouse button, the tab has become its own free-floating window. No big deal, right? I can just plop it anywhere in that array of tabs and it will insert where I put it.
Let’s see what happens when I try that.
Wait, WHAT? Why did my tab move all the way back to the right?
So, despite the fact that tab UI is a 100% solved problem in browsers, Adobe products use this clunky, antiquated, infuriating design. Anyone who has had to manage multiple windows in Illustrator has experienced this masterclass in mediocrity. There is absolutely no excuse for this.
And, unfortunately, Figma is guilty of the same crime. Their desktop app uses the same lousy paradigm in which you cannot insert a tab at an arbitrary point into the array, and when you attempt to drop it in, it just lands on the right. But Figma does two things which make this much less of an issue.
The first is that the tabs are normal height and not the laughably skinny things that Adobe uses, meaning the margin of error at which the tab pops out is much higher.
And the second is that you can just use it in your browser with its infinitely superior tab management. And that’s what I do.
None of that excuses Figma’s oversight, but it does demonstrate the fact that, when you add good UX in one place, it might just protect against bad UX in another.
As for Adobe, they’re hopeless.
Buffer does not warn you that your API token is about to expire
I used to use Buffer to automate my social media posts. Used to, until I got tired of this nonsense:
Occasionally, I would get an e-mail that one of my scheduled LinkedIn posts had failed because the authentication token had expired.
I inquired with Buffer about why this always seemed to happen with LinkedIn and not the other platforms, and they informed me that LinkedIn tokens expired every 60 days. So I asked them the obvious question: if you know the lifespan of a LinkedIn API token, why don’t you inform me when it’s about to expire?
I can hardly imagine a problem with a simpler, cheaper solution. I let their support know that they need to fix this easy-to-fix problem. But, like calendarwork, this problem continued happening every 60 days. I contacted them yet again, asking why they hadn’t fixed it and this was the reply:
Imagine having an easy solution to a major problem staring you in the face and not just taking it because it’s not on the “roadmap”. Here’s a piece of advice, techies: nobody cares about your roadmap except you. Nobody. The rest of us care about a good user experience. That’s what matters, nothing else.
But wait wait wait wait wait, it gets worse:
These peabrains would rather give up years of revenue than implement a fix that probably requires an hour of work at most. It makes my head spin. You best bet I took them up on this offer.
This is what happens when companies protect their incompetent “knowledge” workers with dedicated customer support staff who take all the brunt of customer fury. I guarantee most of these UX problems would disappear if the designers, PMs, and engineers had to handle irate customers instead.
Google Street View randomly changes time periods on you without warning or reason.
This is such an easily avoidable fail and without any discernible upside to Google for allowing it into the product that I’m left without an explanation for how something like this happens.
Imagine this scenario. You are in Google Street View looking at a historical view of a street. As you explore the area, you click a spot directly in front of you, only to suddenly find you’ve traveled through time and whatever it was you were interested in looking at is gone.
Here is an example:
It should be noted that it does not do this because the selected time period is not available in that area. If you move to a slightly different spot, maybe 10–20 feet away, suddenly you can go back to the time period you were exploring.
This means that the software is prioritizing placing the user in something as close to the exact spot down to the millimeter over placing them in more or less the same spot while remaining in the same time period.
Consider that, if someone selects a specific time period while exploring Google Street View, it is because they wanted to see how a given location looked during that time period. They couldn’t give a much more unambiguous signal of their intention because historical images are at least two clicks away, involving a rather obscure button. So, randomly shifting them out of that time period is completely irrational.
But above all, what makes this UX fail so galling is that it’s so utterly unnecessary. As I said above, it is not a result of missing data, and even if it were, Google could just prevent users from navigating to any location without data for that time period. The most mind-boggling part is that I’m pretty sure it would be technically simpler to keep the user in a single time period than have them jump around, which means this design is not only worse, but they put more effort into making it worse.
That’s a shame because, on the whole Google Earth and all of its modules (Maps, 3D View, Street View) are generally well-designed. I usually hold it up as an example of a thoroughly enjoyable, flow-inducing experience, which makes it all the more puzzling that this amateurish blunder would show up.
Spotify’s inability to stick to a consistent one-click song bookmarking model
Here’s a bit of internet history for you. Let’s travel back to the early 2010s. Spotify once had a feature called “Starred”. This allowed you to quickly bookmark songs you wanted to remember for whatever reason. The starred songs then became part of a special playlist. The main thing distinguishing this feature from regular playlists was that you could add songs with a single click, and the playlist had a star icon next to it.
It was a simple, useful feature. But then one day, Spotify decided that users shouldn’t have nice things. They removed the concept of starring and your “Starred” list became a regular playlist. No more one-click playlist for you.
Then, years later, in a tacit admission that it was stupid to remove this feature, Spotify brought back one-click playlists with the “heart” feature.
Hearting behaved pretty much like starring. You clicked the icon and the song was added to a special list with a heart icon next to it. They could have just brought back starring. If you had kept the downgraded “Starred” list all those years like I did, they could have pre-populated your rebooted star list. But that would have made too much sense. No, instead they brought back the concept of one-click playlists but used a different name and iconography.
If the story ended there, it would still merit inclusion in Bad UX Roundup, however, there is a third and almost certainly not final chapter in this Greek tragedy of dumbfuckery. Hearting did not last very long before Spotify decided that it hadn’t imposed enough unrequested changes on users. It replaced the heart with the PLUS.
And if you think that this was just a cosmetic change to the heart button then you really haven’t been paying attention.
Like the original ⭐Starred and ❤️Liked lists, the Plus button has its own special list that you can’t delete or edit, but the Plus button is not readily visible.
To begin with, there is no constantly visible button to add a song to Liked (yeah, the button is Plus but the list is Liked), as you can see below. The green check marks mean that a song has been added to your list and are always visible. But, on desktop you only see the plus button on hover.
And, worse, on mobile, it’s hidden behind a meatballs button:
This pretty much defeats the purpose of the feature since it isn’t one-click liking on mobile, and on desktop version, while it’s technically one-click, the hidden buttons create some sort of Fitts’s law issue while also hindering discoverability.
The entire saga to date reeks of executive hubris.
- First, they insisted that the starred function had to go after they read some fatuous thinkpiece on how Apple is so great because they get rid of features, something something minimalism. Remember, this was at the height of Flat design as well.
- Then, something managed to get through this exec’s tank-like skull which convinced them that one-click playlists are important, BUT if they just brought back the Starred feature, that would be an admission of failure, which a Silicon Valley techbro manchild is incapable of. So, they tried to sneak the feature back in under a different name and because “OMG hearts are like so much more fun than boring stars!”
- Eventually, the fact that it was pretty much the same feature as before began to eat at our hero. They decided they had to make something bigger and better, like it was in the plans all along. Thus appeared the convoluted Plus button. Voila. No longer an organ-shaped admission of failure, but a communion wafer of triumph.
And here I am still just using my downgraded Starred list. It’s not too late.
Serkan, you asked, and you received.
The YouTube comment section doesn’t show what it’s sorting by:
It’s a mystery why YouTube even bothers having a comments section with how little work they actually put into its UX. They must just let the interns work on that part for practice.
Here is what you see at the top of every comment section on YouTube:
It shows you a count of comments and a button that lets you change how the comments are sorted. When you click the button, it shows that you have two options:
As you can see, neither of these options have very long titles, so the question is why don’t they just show what it’s sorted by on the button itself? It should say “Sorted by top comments” or “Sorted by newest comments”. It forces you to make a click to see this very basic information.
If this seems like a kind of weak example to end on, then you aren’t thinking hard enough. The question is why do they omit this information that they could so easily include? It’s because they don’t want you to know how they sort the information by default.
What you see as “top comments” are not really the top comments, because if you look at the upvotes counts, they are not sorted in descending order. If you want to see the all-time greatest comments, you won’t. You’ll see some bullshit arbitrary ranking that nobody understands, not even at YouTube.
This is one more example of Big Tech denying the user the ability to choose what content they want to engage with. Certainly, YouTube excels in pushing what they want you to see rather than what you want to see. This UX fail shows that their attitude extends even to the comments section. After all, you can’t demonetize a comment.
Give me 50 claps if you support good UX
More claps mean more people see my articles. If you want to spread the gospel of good UX design, it’s imperative that you give me the clap(s). Also SUBSCRIBE.
Want more of me?
After you’re done getting your head checked, you can find me at these places.
LinkedIn:
https://www.linkedin.com/in/jclauss/
More articles like this:
http://blackmonolith.co/publications
Barbecue, me and you.
