It’s been a long time since an enterprise software redesign has had an impact as broad as the recent Slack update, and an ever longer time since one has been so universally reviled. Anyone who works in tech knows that redesigns of popular tools – even those that are ultimately successful – face initial resistance and take some getting used to. I’m sure this is what the Slack team is telling themselves right now, as the revolt rages on. But this redesign seems particularly despised, and after using it I don’t like it much myself, so I wanted to try to unpack it. What is it about this massive undertaking that went wrong? And what questions does it raise about technology development in general?
I can’t recall ever seeing The Atlantic run a piece about a UI change, but such was the fervor that accompanied the Slack overhaul. In his analysis, Ian Bogost attributes the changes to Slack’s unquenchable thirst for their users’ attention.
“The interface has always seemed hell-bent on getting you back into the program, even if you’d prefer to do the actual work that your job demands.”
While this is a valid critique of most social media products, I’m skeptical of its application here. In their own blog post announcing the changes, Slack emphasized that the new design would in fact promote focus, and I believe that is what they were trying to achieve. I don’t see how it would benefit Slack, which is now on solid financial footing after being acquired by Salesforce, to lean into its most glaring flaw. If Slack wanted to make the app more engaging, there are a number of tried and true methods they could employ – in-app prompts to turn on more notifications, for example – that would be more effective. In this case, I think Slack did try to make the app less overwhelming and simply failed.
Bogost continues:
“An icon flags unread posts in brightly colored circles. Channel names are bold until you scroll up and down to clear them. Why pick up the phone when you can do an audio “huddle” inside of a DM? Almost all software wants you to look at it, but Slack, a supposed productivity tool meant to help knowledge workers recover from their email, demands more fixation than email ever did.”
None of these things are unique to the new Slack interface, though. “Badges,” as they’re sometimes called, alert users to new messages. In the case of Slack, you can customize which messages result in a badge alert, but by default it happens when you are mentioned in a thread or sent a direct message. Channels with unread content were always bold (as they are in most email inboxes), and I don’t see how audio huddles in Slack are any more distracting than a phone call. These just aren’t the best examples of the very real problem of growth-dependent digital tools using dark patterns to maximize engagement.
So what is wrong with the new Slack? It’s being rolled out in phases over the coming months, and I’ve been able to compare the old and the new, side-by-side, in different workspaces. Here are my thoughts on what Slack was trying to do, but why ultimately the new design is so crazymaking.
The new nav bar adds confusion
The new nav bar is a good example of where Slack’s attempts to simplify went awry. The new Slack design collapses content into Home, DMs, Activity, Later, and an overflow menu that contains Canvases, Files, and many other features.
In theory, this idea of grouping thing to promote focus makes sense: there are fewer things to look at. But having ambiguous groups does not make things clearer or less distracting. Because “activity” is a catchall phrase, I have no idea what the notifications in my Activity section are about. The notification badge on an ambiguously-named section like that simply draws my attention more, since this unread content could be any number of things. In the old Slack, the unread badge on a specific channel gave me more context about the content therein, and I could choose whether or not to engage with it. I will always click into the Activity notifications, because I now need to engage with it to find out what it might pertain to. Not ideal for focused work. The Slack team tried to hedge against this problem by including a preview on hover, to reveal the content beneath, but that itself is a distracting pattern that doesn’t feel clean or intuitive.
Slack seemed to be operating under the general principle that fewer things results in more focus, and I think that premise is faulty. Grouping might backfire if the groups themselves are ambiguously named. When there is less context available on the top level UI, it’s more difficult to safely ignore the content because you don’t know what it relates to.
Doing away with the workspace nav bar
The second issue, interestingly enough, results from the removal of a nav bar but suffers from the same grouping principle in action. In the new UI, Slack got rid of the exposed nav bar that lets users easily toggle between workspaces. This has not gone over well, and Slack fast-followed with a feature that lets you turn this off completely. (The fact that this public rollback and apology – “We’re sorry.” – is pinned to the top of their Twitter(X) page tells you something about the enormity of the blowback.)
The new design groups all workspaces behind the icon of the workspace you’re currently in, in an effort to reduce visual clutter. But in doing this, Slack has again focused on introducing an additional navigation layer and grouping things in a way that is not intuitive or discoverable.
Color schemes and low contrast
The most unnecessary change is also the most profound. Slack is recognized by its signature eggplant brand color. The default color scheme featured this shade of purple heavily, using it as the dark background against which channel names and other text stood out crisply and cleanly.
No more. The new Slack gives you a control panel that allows you to customize your color scheme to a level that rivals a design tool. There are also new gradients, glosses, and other visual polishes that make the new Slack look flashy but also…complex. This seems like over-engineering. Most users will never need this level of customization and granularity, and if anything it presents a choice paradox, forcing people to tweak and refine until they’ve wasted an hour of a Wednesday afternoon finding a Slack theme that feels just right for about 5 minutes before they decide the want to change it again (totally hypothetical example).
The bigger problem, though, is that all of these themes lack the bold contrast that made the original color scheme so appealing. In the new interface, text is harder to read, contrast is lower, and this contributes to an overall fuzzier product that feels less crisp. I think more is less with these color schemes. The old one was beloved for a reason.
Slack seems to be on the defensive after a poorly-received rollout (see above apology tweet). I admire the transparency and willingness to admit fault, but the fact that it came to that says something about how broadly ineffectual this change was. It’ll be interesting to see what future concessions they make to revert things back to the way they were.
So what are the implications of all this? I’m not here to dunk on Slack. This stuff is really hard, and I do think Slack was operating with a good faith desire to improve their product. And I’m sure that we’ll get used to most of these changes eventually. But adaptation is not necessarily the ideal outcome. What if the previous version of Slack was simply better? What do we make of that fact?
When you’re designing for mature products operating at scale, it’s difficult to decide what changes to make, if any. Of course no product is ever perfect, but Slack seemed pretty damn close. I keep thinking about how, if this had been a universally lauded physical object, it would simply exist in its final form, used over and over again, for decades. Software, with some notable exceptions, tends not to work this way. The main reason, I think, is due to incentives within these large organization. Designers come and go, and the new ones want to make their mark. Once a design team exists, it’s saddled with an existential motivation to change things, eventually. Furthermore, new features help companies (even former startups that have been acquired) promote themselves and garner more attention and growth.
It’s unfortunate that change is inevitable. Maybe sometimes a thing is just done. Software will always require maintenance and updates to accommodate new form factors or technological capabilities. But when is enough simply enough? Is there anywhere to go but up?