Skip to content
Kuza Ndoto Academy

Define: Problem Statements

All levelsChapter 3 · Page 9 of 293 min readUpdated 15 Nov 2026

The define mode turns the stories and observations from empathy work into a sharp, focused problem to solve. A well-defined problem makes idea generation far easier, because everyone knows exactly whom they are helping and what need they are meeting.

From notes to insights

After interviews you may have pages of notes. Start by clustering them: write each observation on a sticky note, then group similar notes and name each group. This is sometimes called affinity mapping. Then ask, for each cluster, "What does this tell us that we didn't know?"

An insight is not just a fact; it explains why people behave as they do. "Vendors write debts on paper" is an observation. "Vendors keep debts informal because asking for payment publicly could embarrass loyal customers" is an insight.

Personas

A persona is a short, realistic profile of a typical user, based on your research (not imagination). It helps the team remember whom they are designing for.

ElementExample: "Mama Wairimu"
Who48, sells vegetables at a large Nairobi market, basic smartphone, uses M-Pesa daily
GoalsPay rent and school fees on time; keep loyal customers
FrustrationsForgets who owes what; awkward to ask for repayment
Current behaviourWrites debts on scraps of paper; relies on memory
Quote"If I push for payment, they'll go to the next stall."

Point-of-view statements

The d.school suggests framing the problem as a point-of-view (POV) statement combining a user, a need and an insight:

[User] needs a way to [need] because [insight].

"Market vendors who sell on credit need a way to recover debts without embarrassing customers, because their business depends on loyal, returning buyers."

Notice that the need is written as a verb (to recover debts), not a product (an app). This keeps the door open to many solutions.

"How might we" questions

Teams at IDEO, the d.school and elsewhere turn POV statements into "How might we..." (HMW) questions. "How" assumes a solution exists, "might" says it is fine to explore options that may not work, and "we" makes it a shared challenge.

Too broad How might we fix small businesses in Kenya? Just right How might we help vendors recover credit without embarrassing loyal customers? Too narrow How might we build an app that sends SMS debt reminders every Friday? no clear starting point focused yet open solution already chosen
Figure 1. A good "How might we" question is focused enough to act on but open enough to allow many different solutions.

Generate several HMW questions from one POV, each looking at a different angle:

  • How might we make paying back a debt feel like a favour to the vendor rather than a demand?
  • How might we make debts visible to the vendor without making them visible to others?
  • How might we reward customers who pay on time?
Example

A team of three students at a university in Dodoma interviewed 15 first-year students about settling into campus life. Their clustered notes showed loneliness, confusion about registration and money stress. The strongest insight was that first-years were embarrassed to ask "obvious" questions in public groups. Their POV: "First-year students need a way to get answers to basic questions privately, because they fear looking foolish in front of peers." Their HMW: "How might we make asking a 'silly' question feel safe and normal?" This led to ideas like anonymous Q&A, peer buddies and a "no silly questions" orientation session.

Common mistake

Writing a problem statement that already contains your solution ("Users need an app to..."). Another mistake is building personas from imagination rather than real research. A persona without evidence just repeats your assumptions with a name attached.

Key points

  • Cluster research notes and look for insights that explain why people behave as they do.
  • Personas, based on real research, keep the team focused on actual users.
  • A POV statement combines a user, a need (a verb) and an insight.
  • "How might we" questions should be focused but open to many solutions.
Up next · Design ThinkingIdeate: Generating SolutionsBrainstorming, SCAMPER and choosing ideas to test.
→

Tip: use the ← and → keys to move between pages.