Skipping Over Skeuomorphism

There is a general tendency to opt for skeuomorphic interface design patterns when transitioning to new technology paradigms.

Is it a consequence of our human bias to imagine what is already familiar and established? Or is it justified by strategy to make it easier for users to learn and adopt? Or both?

New tools have a tendency to initially mimic old ones, but they eventually lead to entirely new ways of interacting.

What if new technology did not need to go through skeuomorphic design phases to cross the chasm? What if skeuomorphism was actually slowing down progress, and undermining new technology’s potential?

I am starting to think that skeuomorphism is holding us back. What used to be best practice might no longer be. We might be better off skipping over “mimicking the old”, and going more directly from one paradigm to the next.

Skeuomorphism in this context refers the practice or deriving new interface design from pre-established or familiar UI/UX patterns. For purists, we can also call it “Neo-Skeuomorphism”.

The argument for adopting skeuomorphic design patterns early on is logical and well established. For new technology to be adopted, it first needs to be understood by its users. Skeuomorphic design patterns leverage existing mental models and habits to make it easier for new users to make sense of and start using. Lowering the user learning curve reduces the friction to onboard, accelerates time-to-value, and minimises disruption in the process.

The first graphical user interfaces were filled with skeuomorphic design patterns. A personal computing environment was called a “Desktop”, analogous to our office desks. Data records were organised in “Files” and “Folders”. Internet content was consumed in “Pages”, and saved with “Bookmarks”. In Apple’s MacOS and IOS, skeuomorphism extended to the graphic materials used to represent objects. Wooden shelves in the Newsstand app, yellow lined paper in Notes, and a shiny CD icon for iTunes. Steve Jobs was a major proponent of skeuomorphic design in the early days of Apple. By importing existing references and mental models from the physical world to the digital world, users were more readily able to derive the meaning and function of each personal computing object.

I use the past tense to describe the examples above, but most of them remain in use today. We continue to use the same analogies, despite the fact that some are no longer relevant in the real world.

While skeuomorphic patterns were most likely very important to bridge the first personal computer to the masses, we have significantly evolved since then - both technologically and psychologically - as users, as a market, and as a society.

We might not need Skeuomorphism as much as we used to.

At a time in which interactions with technology were uncommon, and the introduction of new technology infrequent, it made sense to borrow mental models from old paradigms to ease the learning curve and onboard the masses into a new one. In today’s world, the pace of change has gone exponential, and we have become used to it.

We have matured into masters of neuroplasticity, and have developed an ability to adapt and create new mental models faster than ever before. Our cognitive patterns are more flexible, and our behaviours are more plastic. As a result, transitory analogies are less necessary to accompany us through disruption - we are no longer intimidated by new things. Quite the contrary: we are attracted to them.

Avoiding change used to be a protective evolutionary mechanism. For new generations, engaging with change and pursuing it is what it takes to survive and thrive. We grow up exposed to an unprecedented velocity of content and information, and have to learn to filter out signal from noise quickly. As a result, what is old gets classified as “already processed” and dismissed, and what is new is interesting and worth pursuing.

Despite the fact that we know we welcome more change, we continue to use skeuomorphic patterns for new technology. For example, AI interfaces today are in a skeuomorphic phase. We are transitioning away from a “command-based” paradigm of interacting with our personal computers, to an “intent-based” paradigm of interaction. "Prompting" UX is a result a skeuomorphism: we imported command-lines and input boxes to structure our interactions with AI. Yet in an “intent-based” paradigm, input is not necessarily derived from a text box. It is derived from capturing context, and that can take many forms. We intuitively feel that AI prompting is not natural, and yet fail to move-on from it. We continue to force the interaction logic and habits we know into the new interfaces we create, and limit the surface area of what we could experience and use AI for as a result.

Defaulting to skeuomorphic design patterns limits technology’s potential. New technology paradigms require new ways to interact with them, and their mediums of interaction will inform their function and capabilities. Interfaces define how we experience a technology, and what we use it for. Different technologies require different form factors and interfaces. This is precisely why importing skeuomorphic patterns to new interfaces is not a good strategy: new products adopting old interfaces will inherit the old interfaces’ constraints.

The other reason why skeuomorphism fails is that it gets overlooked. What used to be the safe thing to do is now risky: by adopting old design patterns, new products risk being conflated for what already exists, and dismissed faster than they will have the time to prove themselves.

As mentioned above, defaulting to skeuomorphism may not just be motivated by go-to-market strategy, but also a consequence of designer and builder bias. Some design patterns have become so familiar and engrained, that our brains have difficultly imagining other things. There is a shared belief that “we can only create from what we know”, but that belief does not necessarily have to translate into replication. We can imagine new things from combining old and new concepts differently, and in unprecedented ways. Rather than informing our creation from what is already established (i.e. just like LLMs), it is more productive to inform our creation from first principles and imagination. It requires breaking free from what is, and exploring the space of what could be.

I wonder if skeuomorphism is also a manifestation of our risk aversion. We want to avoid disruption, and will subconsciously go for what we know already works. Humans also have an innate desire for social acceptance, which might further reinforce our tendency to favour what is familiar to please the masses and reduce the chance of rejection. These impulses come from our survival instinct, but our survival instinct can be counterproductive to our ambition to innovate.

Perhaps there are better ways to make new technology more accessible that do not limit the design space as much, and perhaps that do not involve design at all. Some strategies to introduce novel interfaces effectively might focus on playing with language to introduce new analogies and create relatable stories. Others might focus more on large-scale content distribution as a means to force visual familiarity through repeated exposure. Onboarding influential early adopters first to subsequently drive acceptance through social proof might also be effective. All of these are marketing tactics, but if well executed and combined, can compensate for the lack of familiarity in design to make new technology more approachable.

Counterintuitively, intentionally diverging from skeuomorphic design patterns can also be an advantage to cross-the-chasm more effectively: once new interface paradigms are introduced and learned, new metal models form, habituation occurs, and the switching cost to revert back to previous design patterns increases. As a result, what is differentiated enough, which is often what is the most bold and new, not only grabs our attention, but retains us as users as well. Breaking free from the cognitive adaptation that occurred over many years is the challenge, but users can intuitively feel if an interface is the right fit for the underlying technology - and intuition is a powerful force.

The opportunity is therefore to ignore skeuomorphism, and liberate new technology from pre-established constraints. To skip over transitory periods, and go straight to the new. To stop accepting the inheritance of tech and design debt, and to give ourselves the license and opportunity to explore new foundations, purpose-built for the new paradigms we aspire to introduce.

I believe we are psychologically trained to welcome new paradigms in raw form, and have the curiosity and humility to experiment with them. I believe that if we overcame our biases, and favoured bold innovation over safe incrementalism, we would surprise ourselves with really interesting things - and the market would welcome them.