Why start with users
User experience (UX) refers to everything a person feels and goes through when using a product: do they find what they are looking for, do they understand what is happening, do they reach their goal without effort? UX design aims to make that experience useful, usable and pleasant.
The most common trap is designing for yourself. The team knows the product inside out, masters the internal vocabulary and imagines uses that are not those of customers. The only way out is to go and see real users, to listen to what they do and what bothers them, before drawing a single screen.
Running example for this course: the fictional cooperative Les Paniers du Coin wants an online ordering app with in-store pickup (click and collect).
Choosing a research method
| Method | What it provides | When to use it |
|---|---|---|
| One-on-one interview | Motivations, context, frustrations (the “why”) | At the start, to understand |
| Field observation | What people really do, not what they say | Physical processes, business tools |
| Survey | Large-scale measurement (the “how many”) | To verify a hypothesis |
| Usage data (statistics, customer support) | Measured real behavior | Product already online |
Qualitative methods (interviews, observation) are carried out with few people, often five to ten per profile, and reveal problems. Quantitative methods measure how frequent they are. You combine them: interviews bring out hypotheses, the survey checks their scale.
Preparing and running an interview
An interview is prepared with a guide: a few themes and follow-up questions, not a form to fill in. The golden rules:
- Open questions: “Tell me about the last time you did your grocery shopping” rather than “Do you do your shopping on Saturdays?”.
- Past facts rather than intentions: what people have done is more reliable than what they think they would do. “Would you use this feature?” almost always gets a polite “yes”.
- No leading questions: “Don’t you think the pickup takes too long?” suggests the answer. Prefer “How did the pickup go?”.
- Don’t present the solution: you are there to understand the problem, not to sell an idea.
- Dig deeper: “Why?”, “Can you show me?”, “What happened next?”.
Take faithful notes (or record, with the person’s explicit consent), separating what the person said from your interpretation.
Synthesizing: the affinity diagram
After the interviews, each observation becomes a note (a sticky note, physical or in a tool such as FigJam or Miro). You then group the notes that talk about the same subject and name each group: this is the affinity diagram.
To estimate the importance of a theme, count the number of participants concerned, not the number of notes: a talkative person who repeats the same complaint three times does not weigh more than three different people.
Theme "Order tracking" 5 participants out of 6
Theme "Pickup time slots" 3 participants out of 6
Theme "Waiting at pickup" 2 participants out of 6Formulating the problem
The research leads to a user-centered problem statement, which will guide the design:
Customers who order from their phone don’t know when their order is ready, so they show up too early and wait in the store.
Then you turn it into an open question that leaves room for several solutions: “How might we let customers know the exact moment their order is ready?” This question does not contain the solution (a notification, a tracking screen, an SMS…): that is the next step.
Key takeaways
- You observe and listen before drawing.
- Qualitative to understand, quantitative to measure.
- Open questions, past facts, no suggestions.
- You count participants per theme and formulate a problem, not a solution.
Exercises
Solutions and tests stay on the server. In the student space, your code is run and checked, and the tutor gives hints before the solution.
Synthesizing interviews by participant
A colleague ran 6 interviews for Les Paniers du Coin and coded each observation by theme. Here are her notes (CSV):
participant,observation,theme P1,"Je ne sais jamais si ma commande est prête avant de partir",Suivi P1,"J'ai appelé le magasin pour savoir où en était ma commande",Suivi P1,"Le créneau de 18 h est toujours complet",Créneaux P2,"Le SMS ne dit pas si tout est prêt",Suivi P2,"J'ai attendu quinze minutes au comptoir",Retrait P3,"Il n'y a que des créneaux en journée",Créneaux P3,"Je rafraîchis la page pour voir si c'est prêt",Suivi P4,"Je ne trouve pas le paiement par titre-restaurant",Paiement P4,"Je dois réserver deux jours avant pour avoir un créneau",Créneaux P5,"Je ne sais pas quand partir de chez moi",Suivi P5,"La file d'attente au retrait est longue",Retrait P6,"Aucune nouvelle entre la commande et le retrait",Suivi- Build the affinity diagram (one group per theme).
- For each theme, note the number of observations and the number of distinct participants.
- Rank the themes by number of participants.
- Write the main problem statement and its “How might we…?” question.
- The 12 observations are sorted into 4 named groups
- The count distinguishes observations from distinct participants
- The ranking is based on the number of participants
- The "How might we...?" question does not contain a specific technical solution
Rewriting interview questions
Here are four questions prepared by a colleague for interviews about in-store pickup:
- “Would you like to receive a notification when your order is ready?”
- “Don’t you think the wait at the counter is too long?”
- “Do you do your shopping on Saturdays?”
- “How many times would you use an automatic pickup locker?”
Rewrite each of them as an open question about a lived event, without suggesting an answer or presenting a solution. For each one, state the flaw you fixed.
- Each rewritten question is open (it cannot be answered with yes or no)
- No rewritten question mentions a feature (notification, locker)
- The questions are about a past experience ("the last time...")
- The flaw of each original question is named
Which method for which question?
The team wants to know why customers abandon their order at the time slot selection step. Which method provides the most information about the “why”?
- A survey sent to 2,000 customers with checkboxes
- One-on-one interviews with customers who recently abandoned at this step
- The abandonment rate calculated in the analytics tool
- A team vote on the most likely cause
End-of-lesson quiz
Which of these interview questions is best worded?
- "Would you use an app to place orders?"
- "Tell me about the last time you ordered your groceries."
- "You surely find the current site complicated, don't you?"
In a synthesis, one participant mentions the same problem three times. How do you count it?
- Three times
- Once, as a number of participants
- It is excluded from the synthesis
Which wording is a good “How might we…?” question?
- How might we add push notifications?
- How might we help customers know when to come and pick up their order?
- How might we redo the site?
The rest of the course
- Personas and User Journeys17 min
- Information Architecture and Wireframes17 min
- Mockups and Prototypes in Figma18 min
- Testing with Users and Iterating18 min