How human-centered design helps teams navigate uncertainty

Executive overview

Most teams are good at solving problems but poor at finding the right ones. Human-centered design is a structured approach to closing that gap — tuning into what people actually need before building solutions.

The d.school's toolkit spans team empathy practices, assumption-challenging exercises, and AI-augmented scenario planning. Together they help organisations move faster, learn more from failure, and avoid building the wrong thing.

The core insight: design's primary value is not aesthetic — it's epistemic. It helps you figure out what problem is actually worth solving.

What human-centered design is

• Everything not found in nature has human influence — design is unavoidably present in all of it
• Human-centered design focuses specifically on tuning into what people need, considering natural and social context, then generating solutions
• The d.school specialises in creative problem-solving under conditions where there is no predetermined right answer
• It incorporates graphic design elements but extends to team structure, space design, and organizational behavior

Building connection in distributed teams

• The wordless conversation exercise: partners exchange hourly photos throughout a day, building empathy without shared language or prior relationship
• Physical prototyping matters even remotely — having the same materials in front of dispersed team members creates shared experience
• Building to think: physically drawing, collaging, or whiteboarding creates different cognitive sensitivity than screen-based work
• Analog materials leave more open to interpretation and surface misunderstandings early
• Remote shadowing (via Zoom) can work as well as in-person — sitting with a colleague from start to end of their day

Organizational design as companies scale

• Early-stage teams in one room develop high natural empathy for colleagues' roles and challenges
• The moment teams split across rooms or offices, that empathy degrades rapidly — knowing this horizon is coming matters
• Internal conflict often signals teams have gathered different data without synthesizing it together — fix by creating shared raw-data review sessions
• Apply customer-research practices (like shadowing) internally to understand how colleagues experience their own work

Assumption storming

• Brainstorming is well known; assumption storming (developed by Craig Lauchner) is less common but highly effective
• Collect expert opinions on a space, then list all the assumptions embedded in them
• Categorize assumptions as facts, opinions, or guesses
• Challenge the opinion- and guess-category assumptions: "If this were not true, what would we design?"
• Useful for de-risking launches, challenging outdated mental models, and opening creative space

AI as a design tool

• AI is most useful in fast flaring — rapidly broadening the aperture of possibilities before a human focuses and evaluates
• Prompting AI to "think like a futurist" and return 5–10 divergent scenarios generates more creative starting points than solo ideation
• The d.school is building a generative case-study tool: teams are placed in future scenarios, interview simulated inhabitants, develop solutions, then encounter a deliberate tilt (failure event) that forces adaptation
• AI-assisted rapid prototyping ("vibe coding") lets student teams break internal stalemates by building working prototypes quickly rather than debating abstractions
• AI makes labor-intensive reflective tools more scalable — tools that were too time-intensive to use consistently can now be embedded in regular practice

Productive struggle and the learning journey map

• Learning journey maps plot two lines over a project's timeline: how much you were learning, and how you were feeling
• Where those lines diverge — high learning, low affect — marks productive struggle
• Struggling during skill acquisition produces more durable learning than easy progress
• For teams: use the map at project midpoints or end-of-cycle reviews, not just as evaluation but as learning infrastructure
• Framing retrospectives around "what can we learn from this?" rather than "what went wrong?" lowers defensiveness and increases candor

Problem finding before problem solving

• Haley King's startup Paxos Appeals illustrates the difference between problem solving and problem finding
• Starting premise: 10–20% of health insurance claims are denied; only 1% of people appeal, yet 40% of those who appeal succeed
• Initial assumption: the problem was increasing awareness of appeals
• What design revealed: the most critical pain point is the act of writing the appeal itself — people need support in that specific moment
• Refocusing on that lever led to proprietary software comparing policy terms, medical documentation, and generating appeal drafts with a human in the loop
• A further human insight emerged: people value the feeling of having someone on their side, even independent of financial outcome
• Early-stage startups cannot spread resources thin — design helps identify the highest-leverage problem to start with