Beyond UX surveys and heatmaps: how to design for latent user needs in Southeast Asia's mobile-first markets before users can articulate them.
The most honest thing a user has ever told you is probably incomplete. Not because they’re hiding something — but because they genuinely don’t have the words for what they want yet.
UX Collective contributor Hiroshi Sato frames this through the Japanese concept of the engawa — the transitional veranda space between the interior of a house and the garden outside. It’s neither fully inside nor outside. It’s where you sit when you don’t quite know what you need. Sato’s provocation: the best design lives there too, in the threshold between what users can articulate and what they actually desire.
For digital teams in Southeast Asia, this isn’t poetic abstraction. It’s a data problem with a design solution.
What Behavioural Data Reveals That Surveys Don’t
When I look at clickstream data across e-commerce flows in the region, the most instructive signals aren’t conversions — they’re the pauses. The 4.2-second hover on a product image before a bounce. The abandoned cart that started with three size changes. These micro-hesitations are the engawa moments: users in-between a felt need and a formed decision.
Shopee’s search-to-purchase funnel in Indonesia offers a useful case. Their internal UX research (shared at a 2024 regional summit) found that users who used the visual search feature converted at 1.8x the rate of text searchers — not because visual search is inherently better, but because it lets people articulate needs they couldn’t type. The interface met them at the threshold.
The tactical implication: instrument your analytics to track hesitation, not just action. Session replay tools like Hotjar or Microsoft Clarity, cross-referenced with segment-level cohort data, can surface where interface friction is actually masking unarticulated desire.
The Cultural Dimension Southeast Asian Teams Often Miss
Designing for latent needs isn’t culturally neutral. In high-context communication cultures — and much of Southeast Asia qualifies — users are less likely to explicitly state dissatisfaction or confusion. They’ll simply leave. That 68% bounce rate on your Thai-language landing page isn’t indifference; it’s unexpressed expectation.
This creates a specific design challenge: the interface has to do more interpretive work than Western UX frameworks typically assume. LINE’s Thai interface evolved its sticker-selection UX precisely because qualitative research revealed that emotional expression through imagery was a primary communication mode — not supplementary. The product team didn’t ask users “do you want more stickers?” They observed what users reached for when words felt insufficient.
For multilingual interfaces specifically, this matters at the component level. A CTA that reads confidently in English can feel abrupt or even rude when translated directly into Vietnamese or Bahasa. The button copy is the least of it — the surrounding whitespace, the button size, the proximity to other elements all carry meaning that shifts across linguistic and cultural contexts.
When Design Protects Its Own Value (And Why That’s a UX Issue Too)
Illustrator Raymond Biesinger recently published a self-defence guide for creatives based on nine documented instances of being underpaid or exploited, as covered by It’s Nice That. The book is partly cathartic, partly practical — but it surfaces something the UX community rarely discusses openly: the design brief itself is a designed artifact, and poorly structured briefs produce poorly understood work.
This matters for in-house digital teams in Southeast Asia managing agency and freelance relationships. When the design deliverable is vague — “make it feel premium” or “more Southeast Asian” — you’re not giving designers room to solve a real problem. You’re giving them room to guess, which produces work that satisfies no latent need because the need was never surfaced in the first place.
The fix is structural. Before a design brief goes out, run it through a simple test: can you connect every design requirement back to a specific user behaviour or business outcome? “Increase visual hierarchy on the product page” is actionable. “Make it pop” is a wish the designer cannot yet wish alone.
Biesinger’s frustration with exploitative clients has a mirror image in frustrated clients who receive work they can’t use — both failures trace back to the same broken translation layer between intent and execution.
Making the Invisible Legible — Operationally
The practical challenge of designing for unarticulated needs is that it requires a different research posture. Standard A/B testing tells you which variant performs better among users who already understand their options. It can’t tell you what option you’re missing entirely.
Three approaches that work in Southeast Asian market conditions:
Longitudinal session analysis over point-in-time testing. Track how the same user segment interacts with your interface across 30-day windows. Desire evolves — and the pattern of evolution is often more informative than any single session.
Contextual intercept research in-app. Grab-level NPS is a model worth borrowing: short, triggered immediately after a specific interaction, asking about that interaction only. Not “how was your experience?” but “what were you looking for just now that you didn’t find?”
Cross-platform signal triangulation. If a user abandons on web but converts on app for the same product category, the delta is interface-driven, not intent-driven. That gap is a design brief.
The engawa is a useful metaphor precisely because it refuses to force a binary. Good design in Southeast Asia’s complex, mobile-first, multilingual markets has to hold that same ambiguity — meeting users at the threshold, not demanding they arrive fully formed.
The question worth sitting with: if your current design process can only respond to what users explicitly request, what category of desire are you systematically building blind spots around?
At grzzly, we work with digital teams across Southeast Asia to connect behavioural data to design decisions — turning the signals hidden in your analytics into briefs that designers can actually act on. If your UX process feels like it’s optimising a journey your users have already outgrown, that’s a conversation worth having. Let’s talk
Sources
Written by
Mellow GrizzlyTranslating raw data into activated audience segments, predictive models, and decisioning logic. Comfortable at the intersection of the data warehouse and the campaign manager.