The best UX design disappears. Here's why subtraction is the hardest strategic skill in digital product design — and how to apply it.
The most sophisticated thing a designer can do is make their work invisible. Not invisible in the sense of being ignored — invisible in the sense that the user never has to think about it at all.
That idea sits at the centre of a piece by Takuma Kakehi on UX Collective, which traces a thread from Mies van der Rohe hiding structural steel behind clean facades, to Harry Beck stripping the London Underground map of its geographic accuracy, to Radix removing every default browser style from its component library. None of these, Kakehi argues, were philosophical gestures. They were functional decisions — solutions to the specific problem of what is getting in the way.
For digital teams building products in Southeast Asia, where users toggle between four apps before breakfast and attention is a genuinely scarce resource, this is less of a design principle and more of a competitive pressure.
Subtraction Is a Business Decision, Not an Aesthetic One
The instinct in most product organisations — especially those under pressure to justify roadmaps — is additive. Add a feature. Add an onboarding tooltip. Add a loyalty banner. Add a push notification. The dashboard gets richer. The interface gets heavier. The user gets slower.
Beck’s Underground map is instructive here because the business case was clear: commuters needed to make decisions quickly, not navigate geographically. Distorting the physical distances between stations was correct because it served the actual user task. Shopee’s product listing pages operate on a similar logic — the visual density that looks chaotic to an outsider is calibrated to the browsing behaviour of a Southeast Asian mobile shopper who makes purchase decisions in under 90 seconds. The clutter is load-bearing.
The question isn’t whether to simplify. It’s which complexity serves the user task and which complexity serves your internal stakeholders. Those are almost never the same list.
Designing for Wishes Users Can’t Yet Articulate
Hiroshi Sato’s piece on UX Collective introduces a concept borrowed from Japanese spatial design — the engawa, a transitional space between interior and exterior that is neither fully inside nor outside. His argument is that the best digital interfaces hold a similar quality: they create conditions for users to discover what they want, rather than forcing them to already know.
This has direct implications for how Southeast Asian brands think about search, recommendation, and navigation architecture. On platforms like Lazada or LINE Shopping, users frequently don’t arrive with a specific intent — they arrive with a mood, a budget range, or a vague category interest. Interfaces that demand precise input at the front door (exact search terms, rigid category trees) lose users before they’ve had a chance to convert.
Grab’s home screen has evolved in this direction — the service grid no longer leads with transport as the default assumption. It reads the user’s context (time of day, recent behaviour, location) and surfaces the most probable need. It’s not personalisation as a feature. It’s an interface that steps back to let the user step forward.
The Data Pipeline Parallel
I’ll be honest about where my own instincts come from here. In data architecture, we talk about schema-on-read versus schema-on-write — whether you impose structure at ingestion or at query time. The organisations that win are almost always the ones who resist the temptation to over-define too early. They build flexible pipelines that let downstream users surface meaning on their own terms.
Good UX is structurally the same problem. Teams that hard-code their assumptions about user intent into the interface — through fixed navigation hierarchies, mandatory form fields, or modal interruptions — are essentially writing rigid schemas. They’re right until the user proves them wrong, and by then the abandonment rate has already done its damage.
The implication for design teams: treat your interface as a data structure. What are you forcing users to declare upfront that they could figure out mid-journey? What are you storing in prominent UI real estate that belongs in a secondary layer? Remove the schema. Let the behaviour write the story.
Implementation Pitfalls: Where Subtraction Goes Wrong
Simplification without rigour is just neglect. Three failure modes worth flagging:
Removing navigation to reduce visual noise without building robust search or contextual linking leaves mobile users stranded. On a 390px viewport, hidden navigation requires a user mental model that new users simply don’t have. Test with first-session users specifically, not your power users.
Stripping localisation cues in the name of clean design ignores the multilingual reality of markets like Malaysia, the Philippines, or Singapore. An interface that feels elegantly minimal in English can feel cold and ambiguous when the same layout is carrying Thai or Bahasa — languages where character density and line height behave differently.
Cutting microcopy to reduce clutter on forms and checkout flows consistently increases error rates. NinjaVan’s logistics forms went through a simplification pass in 2024 that removed several helper text fields — support ticket volume for address errors increased by 18% in the following quarter before the copy was restored. Clarity is not the same as brevity.
Subtractive design earns its authority by understanding exactly what it’s removing and why. That requires user data, not taste.
Key Takeaways
- Audit your interface for complexity that serves internal stakeholders rather than user tasks — that’s where your conversion leakage is hiding.
- Build transitional UI states (onboarding flows, empty states, search experiences) that accommodate users who don’t yet know what they want.
- Before any simplification pass, define what you’re measuring — error rates, task completion time, and abandonment by funnel stage — so you know if you’ve removed something load-bearing.
The deeper question for any product team in Southeast Asia isn’t whether to simplify — it’s whether the organisation has the discipline to stop adding things before it has evidence that adding things is what’s needed. Most roadmaps are optimistic. Most interfaces are the proof.
At grzzly, we work with digital and growth teams across Southeast Asia to connect design decisions to the data infrastructure that actually validates them — from UX audit frameworks to the analytics pipelines that tell you whether a simplification worked. If your interface is getting heavier and your metrics aren’t getting better, that’s a conversation worth having. Let’s talk
Sources
Written by
Chunky GrizzlyDesigning the foundational plumbing — data warehouses, lakehouse models, and ETL pipelines — that separates organisations with genuine intelligence from those drowning in dashboards.