Quality The Business of Software

Don’t piss off your users by suddenly changing your UI

When you dump a whole new UI on your users, expect that they will hate it and you.

By Jim Grey (about)

Delivering software on the Web is great. Especially with continuous delivery, we can deliver changes large and small anytime we want. And then we can get quick feedback from our users and the market, adjust the software accordingly, and push those updates fast, too. It’s utopia and the Holy Grail rolled into one!

Except that users are not very excited when we change things. They want software to stay as it is. Well, mostly: they want us to fix the bugs that affect them, of course, or even to add this or that little feature. But please, they plead, don’t make it work differently than it does now.

Meanwhile, we face various pressures. Markets shift; new needs emerge and old needs become less important. Technologies shift; old frameworks become outdated, new frameworks enable us to keep pace. Today everything has to not only work on mobile, but feel native to mobile — and all run on a single codebase. This is shifting product direction across our industry.

That’s the backdrop against which WordPress, the content engine behind one out of every four Web sites, rolled out a new editor last week. It was part of a complete rewrite of all of Their old technologies just couldn’t stretch to where the world was moving. So they threw it out and started from scratch. Their new editor is fast — fast! — and works fluidly, while looking great both in my browser and on my phone.

Spanking new editor in my browser…

But boy, were users pissed. Pissed! Check the forum: 19 pages of complaints and counting. Sometimes, I swear, users wouldn’t be happy if you sent them gold bars, because they preferred the silver bars you used to send them. But very often users have a point: they’ve gotten into the swing of your software, and now you’ve changed it and they have to learn it all again. Worse, maybe now something they used to be able to do isn’t there anymore, or if it is, they can’t find it.

…and on my phone

For the record, I was the first commenter on that thread, because I experienced some of those frustrations. I tried to be kind, but several features I use either went missing or were accessed in a way that I couldn’t easily discover. Argh! And I wished the editing space were wider; it felt awfully cramped. I wasn’t alone in any of these complaints.

I wanted to just edit a post. I didn’t want to learn a new interface. But I found that there’s no way to just revert to the previous editor. It is simply gone.

I understand what drives changes like this and know that this is a monumental achievement this is for WordPress. Still, because I’m a heavy WordPress user, more than anything else I feel frustrated. The new editor breaks all of my usage flows, and I’m having to rediscover everything. I didn’t want this.

It’s the same, by the way, with Microsoft Office’s ribbon, which replaced an older menu structure way back in Office 2007. That’s forever ago in software terms. Yet there are still features I can tell you exactly how to find in those old menus, but I have to Google where they are on the ribbon.

Users don’t give a rip about your business or the future of technology. They use your product to accomplish a thing. As long as they can consistently and easily accomplish that thing, they stay happy. Many users learn your product’s nuances and become quite adept with them. When you suddenly change the UI and all of their flows are interrupted, of course they’re frustrated.

So what would happen if you followed Basecamp’s model? Their software helps companies small and large manage projects. Last month, they released Basecamp 3, a ground-up rewrite — yet they received not a single complaint from existing users. That’s because Basecamp 2, and for that matter Basecamp 1, remain fully active. Existing users can upgrade if they want, or stay put if they don’t. There are compelling reasons to move to Basecamp 3. But if you’re a happy Basecamp 1 or 2 user, those products will be there, fully supported, for as long as you want to use them.

Maybe your company can’t do that. But what can you do to ease the transition for your users, so they can stay fully productive? Think this through. It’s more important than any technology or implementation decision you make.

Fortunately, WordPress does, for some reason, still provide back-door access to an even older editor.

Outdated but highly functional classic editor

I don’t care that this is based on outdated technology: it’s fully featured, and I know how to make it sing. I cut my blogging teeth on this editor when I started my personal blog in 2007. I’ve written over a thousand posts in it. I hope it never goes away.

By Jim Grey

Writer. Photographer. Leader of geeks.

9 replies on “Don’t piss off your users by suddenly changing your UI”

Hear, hear! Like you, I’ve been gnashing my teeth trying to figure out how to accomplish simple tasks I used to do automatically. (And I have the same complaint about Apple, whose software changes occur even more often and seem even more capricious.) You make an excellent point in stating that WordPress would do well to copy Basecamp and leave older versions running for a while. Or, if that’s not feasible for some reason (server space? I dunno) at least let us know well ahead of time that a big change is coming and give us a tour of what’s new. Sigh.


I think we’re in a time where the industry is figuring out how to use the new tools and methodologies (i.e., continuous delivery) to their fullest advantage, yet help users through the transitions. I’m trying to be a voice here who strongly encourages paying attention to that transition. Basecamp’s model is interesting but has associated costs that they’ve had to figure in. Flickr has rolled out changes slowly, and their current UI is a franken-UI of old and new stuff that is transitioning slowly into their future. (That hasn’t helped Flickr users not lose their bloody minds at every change.) When Google rolls out UI changes, they usually have popups that walk you through.

UI changes are inevitable. We in software development have to think harder about how to help our users cope.

Liked by 1 person

Web-based software has the opportunity and means to move faster, given the company driving the change is committed to making an evolutionary jump (vs a split). I feel a great deal of empathy over your frustration with change, but I also agree with your argument that the software has to change—so it’s an interesting paradox. Better transition is a missed opportunity here.

Can’t help but mention 🙂 that the feedback thread you referenced is not entirely made up of complaints. At least one third of replies there are from happiness engineers helping people, some are simply help requests or unrelated, some actually weren’t related to the change at all (this happens often), and there are a small number of kudos in the mix too.

Cheers Jim. I love your blog!


Sheri, I’m both glad to see you comment and a little embarrassed as I didn’t want my words to hurt. I kind of forgot a couple Automatticians are in my audience. If someone wrote a post like this about my company I’d see the point but it would still sting. So, I’m sorry!

I have an axe to grind that really has nothing to do with WordPress in particular; this was just an example I could use. My axe is this: we simply must do better in our industry at helping our users transition as we evolve our products. I’ve also used Flickr to grind this axe:

And I’ve been guilty of perpetrating this kind of change. At my last company, we rewrote our product ground up and released it. We did do some things trying to help our customers transition, but it turned out not to be nearly enough. I learned some valuable lessons from that experience.

You’re right: the comment thread is heavily populated by happiness engineers carefully responding to every issue raised. Thank you for shining a light on that, and thank you for suffering the slings and arrows!

Liked by 2 people

I wasn’t hurt! I love this article, and the 2nd part of my comment was just because I wanted to add a small note to make sure to celebrate what I think is a success in that thread: constant help for users and happiness engineers working closely with them through the transition. I think it’s a first, basic step even—and something happiness engineers can do to help with transition where they cannot otherwise (i.e. they are dependent on developers and decision-makers for more proactive types of transition).

I am so with you about the valuable lessons learned component for these experiences. Super interesting topic!

Liked by 1 person

Anything can be learned, even something that’s poorly designed. It’s harder to unlearn, however! That’s the real issue for me, an Office old timer and former power user.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.