Users Don’t Hate Change, They Hate Losing Habits
I have been asked many times to "do a complete redesign of the product". An ordinary task, one would say, every product should be redesigned from time to time, then why not? Especially if there are known issues and obvious constraints dictated by current design. "The product is crying out for a redesign", I've been said; let's do that!
Except for a small detail: the product in question has a big user base and works very well for its users. Product testimonials are all positive, praising its usefulness and practicality. Business metrics are growing quarter by quarter. In a word, it's successful.
But the business stakeholders are nevertheless asking for a redesign. Why?
Rationales for an (un)reasonable change
To find out about the reasons for the everlasting need for product change, we should analyze those who request a redesign - decision makers. In my experience, many of the business stakeholders and product managers are change-driven and impulsive. Understandably so, because it is their job to manage other people and projects. They easily get bored with the existing version of the product because they spend most of the time looking at it, working on updates. They also analyse competitive products; trying to catch-up with others could create the illusion of falling behind.
But boredom or panic is not a design strategy. Those are two real human emotions that are hard to control, and product stakeholders frequently resort to default command mode: let designers do the redesign! Let them propose "new version", update the aesthetics, make it more glamorous, apply popular design trends to our product; it will get us out from boredom, it will be more like other successful competitive products, it will sell better!
Well, this is not true. A redesign without a real need, done for its own sake and focused on surface only - is doomed from the start. Let's see its classical case.
One catastrophic redesign failure
Windows 8 launch in 2012. is a typical case of a redesign fail. The Microsoft executive team decided to remove its iconic "Start" button and menu.
This design decision hit hard on loyal users. Their response was predictably negative; the new version required users to change their long-established habit for the essential software they use every day: the operating system.
Also, some core functions like Search, Share, Settings, Devices - once at obvious place, were hidden; therefore users lost the way to access them instantly, as they were doing before. There were no affordances, nor discoverability for those crucial features.
The feature used every day - "Shutdown" - was moved to a different part of the interface. "How do I turn off Windows 8" become the most frequent search query back in the day.
And why the change? Microsoft noted that Start button usage was declining and used it as a rationale for the change. But "low usage" didn't mean "low value". They wrongly concluded that Start button was becoming redundant.
There was one stronger reason for the change: Microsoft was not catching up in mobile market. By redesigning Windows to prioritize mobile users, Microsoft hoped that developers would extend app development from desktops towards mobile devices - tablets and phones. They wanted to use desktop users as a leverage towards mobile market. This didn't happen.
Some of those changes were reverted back in next Windows 8.1 update, after a year of a user backclash. Until then, users downloaded the extensions that brought their loved "Start" button 10s of milions of times.
There are many reasons for failures of redesign project. Here are the ones that do the most damage.
Weak criterion - or absence of it
When a design team has no goal beyond "making it new by redesigning," it has no clear definition of "done". It has no clear success criteria for the project, either[1]. And if there is no stated problem in the redesign project then the surface change becomes the single measure of progress. Consequently, opinions and preferences of decisions-makers take precedence over the real business metrics.
So we are then left with the single criteria: how well the product looks. This is because people believe that visually attractive products are easier to use. It is a well-documented human universal, called the aesthetic-usability effect. Because of this bias, focusing on appearance over substance can easily hide usability issues. If there is no real criterion for redesign, the product people easily slip into this bias.
The surface redesign changes a perception. If done properly, the product that looks fresh can attract new user base and expand to new markets. But after a first test, the product made only to be visually beautiful quickly reveals its structural and usability flaws.
Despite its large potential, it might be a trap for design teams if "to make it look new" is the only reason for redesign. And there lies the next common cause for unreasonable redesign.
"New is better" paradox
"Make it new" is not a design criterion - unless you work in the fashion industry. People mistake "new" for "better" - hence the confusion. Honestly, "Make it better" is a sensible goal, meaning that the product team is about to deliver value.
But "new" is not the same as "better". New is new, and better is better. This important distinction is forgotten, more often than not. And from there, somehow, developed "new tech is better than old tech" - a popular yet shallow mantra of technological progress. "Throw out the old, bring in the new!" - add some features, polish the visuals or rebrand, stamp "UPDATED RELEASE" on it and sell it!
Actually, "make it new" was a motto for modernism in 30s', persuading the clientele that novelty is a value in itself. Then the public perception took over, and sculpted a mental model of "better" around novelty. This automatically put "old" in an inferior position.
Modern digital products often treat familiarity as a problem to solve, directly sabotaging the evergreen principle of familiarity. Especially for consumer technology, novelty can be part of the appeal[2].
So: new - good; old - bad. Then "make it new", disguised as "make it better", came to mean: just do a redesign and the product will automatically get better. Which brings us to the next temptation - possibly the most dangerous of all: doing for doing's sake.
Doism
To the man with the hammer, everything looks like a nail
For design fundamentalists, "I design because I can" becomes "I design, therefore I am". The designers design, and redesign, because they can. The digital medium and tools allow them to do that easily. Making things just because you can, for no good reason has even been given a name: “inverse vandalism”[3].
Alas, "doism" ends badly most of the time. This is because of a lack of rational and critical thinking in the moment, especially under time pressure.
"Better than nothing" is worse than "nothing".
Generative LLM technologies[4] force a "better than nothing" approach. They are perfect for that. Building products with them relies on confirmation bias and precisely because of that produces subpar results.
And if the "doism" of designer is not enough, a reminder: it's the product manager's job to keep the team busy. When the backlog is empty, it's tempting to make a product feel "new" by redesigning it. Doing a redesign also gives the illusion of progress, especially to unconfident product managers.
Then when it's the right time for a redesign? Should it ever be done? Yes of course - Lets see!
When a redesign makes sense
Some legitimate reasons for redesign are:
- Design or technical debt
- Change in business rules
- Technology forces a redesign
- When the old design limits existing users or doesn’t work anymore
- Expansion to new markets and new user base
The reasons for redesign should be evidence-based. They should have their roots in user observations, usage analytics, measurable business metrics or feedback. And those reasons are almost never strong enough to justify a total redesign.
Apart from specialized industries that sell "new" as an advantage of a product, the need for a "complete design overhaul" is rarely driven by a genuine product need. This is most evident in well-established products with an active user base. The users want familiar design.
However, the mantra “if it works, then don’t fix it” doesn’t always work. Designer's deceptively convincing excuse is consistency: keeping the same interface for years builds habits in users, raises perceived usability, and makes the product easier to use over time. Users become familiar with the product regardless of its usability: they adapt to wicked interfaces or non-optimal user flows; they develop odd shortcuts to avoid obstacles in their daily work (I’ve seen this many times).
What should designers do when asked for a thorough redesign? They should fight back with "why". This is the universal question, aimed at a productive conversation: talk about the reasons for the redesign. Look for the root cause. If there is none, or it's weak, try to get the decision reversed and propose incremental changes instead. Remember, the product manager might be bored with the current product, but customers most likely are not.
Once all stakeholders understand the real reason for the redesign, the team should set a clear goal for what the redesign should achieve.
If everyone agrees there are clear rationales for a total redesign, designers should still perform it incrementally. Do it carefully, with a plan, in phases. Then, during the rollout, measure the key metrics carefully and change the project direction if needed. For example, if the user feedback is strictly negative in the early phase, re-evaluate the decision to do a total redesign. Sticking to a plan that does not deliver as expected will hurt the existing business. Nobody wants that.
Will this approach lead to a “local maximum” of design? Yes. It can prevent a “moonshot”. And that is the point: be cautious not to disrupt user habits when performing a redesign project; the last thing you need are mad users.
But if the clear goal is a "moonshot", with its known downsides, then go for it. Wipe out the old and do a complete redesign. Even then, you should make a plan before redesigning and offer existing users the option to revert in case the redesign doesn't work.
Today's products are viewed as "innovative" or "different" just because they are "re-designed"; as if "being redesigned" were the only quality. But this misses the point: successful redesign is supposed not to be substantial. The redesign should lean toward the functional aspect of the product, toward "tweaks" and fixes, rather than "complete overhaul". Not every redesign should be revolutionary, nor dramatic. For example, Wikipedia 2022 redesign is barely noticeable, and that’s the point.
Designers should not avoid a redesign at all costs - it is worth doing under certain circumstances. In general, it should be thoroughly considered and done under controlled conditions.
The Evergreen Principle: Respect the Habit
When you redesign, you force users to change their habits. After launch, it might seem that people hate the redesigned product; in fact, they hate losing their habits. Once a habit is established, people would rather not change it.
“You know… we, the users, don’t like it when you change things in the app. We get used to it. Then you change it and we can’t find anything. We don’t want you to change it.”
Redesigning for the sake of it will not give the product a market advantage. A functional design doesn't need a refresh or a redesign unless the requirements change dramatically[6].
There is nothing wrong with being "boring" when designing. Boring can be a standout feature and an advantage, especially for established users who care more about usability than aesthetics.
Ask yourself: does the redesign add meaning or solve some of the crucial problems? If so, change carefully then test and measure the impact. Make people's habits your first priority.
Change with care, not with ego. Solve problems and add value - anything else is just redecorating, refreshing or restyling.
In the case of Windows 8 redesign, though, Microsoft's executive team knew exactly what they were doing: they deliberately wanted to steer the platform away from the desktop paradigm and toward to the mobile one by favoring the touch interface. They clearly had the problem they needed to solve. But this came at great cost to existing users - the fix cost them the desktop and never won them the mobile. ↩︎
https://deadparrotbbs.com/why-old-software-sometimes-feels-better/ ↩︎
Coined by Alan Kay ↩︎
Incorrectly called "AI" ↩︎