Day 2: Ideas and testing
Go wide with ideas, then let real people tell you which ones matter.
Today's goal
By the end of today you can generate many ideas, choose one with a clear method and design a cheap experiment to test it.
Agenda
- Warm-upYes, let's10 min
- LearnHow to ideate well15 min
- ActivityFifty ideas in twenty minutes30 min
- ActivityChoosing one idea20 min
- BreakBreak15 min
- LearnYour idea is a set of guesses15 min
- ActivityCustomer conversations35 min
- ChallengeDesign a no-money experiment25 min
- ReflectEvidence or opinion?10 min
- QuizDay 2 quiz5 min
The session
Yes, let's
A d.school warm-up that trains you to build on ideas instead of blocking them. This version is seated and low-contact.
- One person suggests an imaginary activity for a group, for example 'Let's open a juice stand on the moon.'
- Everyone answers 'Yes, let's' and the next person adds a detail: 'Yes, let's, and serve it in floating cups.'
- Keep going for 2 minutes per scenario, then start a new one.
- Notice how quickly ideas grow when nobody says 'but'.
Solo
Write a chain of 10 'Yes, let's, and...' additions for two silly scenarios.
Group
Play in circles of 6 to 8, seated. Anyone can pass.
How to ideate well
The d.school treats ideation as going wide before going narrow. These are its key points.
- Separate generating ideas from judging them. In a brainstorm, judgement does not come into the discussion.
- Capture every idea, whatever you think of it.
- Build on other people's ideas and listen to each other.
- Go for volume: the Bootleg suggests shooting for 50 ideas in 20 minutes.
- Start from a 'How might we' question, not from a product you already like.
- Watch for bias: falling in love with your first idea is a form of anchoring (see the critical thinking lessons).
Related lessons: Cognitive Biases: Why Smart Brains Make Mistakes, Thinking like an entrepreneur: spotting problems worth solving
Fifty ideas in twenty minutes
A timed brainstorm on your best How might we question from Day 1.
- Write your How might we question at the top of a large sheet.
- Set a 20-minute timer. Write one idea per sticky note, with a short headline.
- Every 5 minutes, push in a new direction: a free idea, a digital idea, a physical idea, a service done by hand, an idea for a much younger or older user.
- Count your notes when the timer ends.
- Spend the last 10 minutes grouping similar ideas.
- Produce: at least 30 ideas, ideally 50, grouped into clusters.
Solo
Work alone with a timer, or invite one family member for the last 10 minutes to add ideas using 'Yes, and'.
Group
Teams stand or sit around one sheet. The facilitator (nicely) enforces the rules: no 'but', no judging, all ideas captured.
Choosing one idea
The d.school warns not to narrow too fast, then gives simple ways to choose.
- Post-it voting: each person gets three votes and marks three ideas they are drawn to.
- Sort the top ideas into the d.school's four categories: the rational choice, the most likely to delight, the darling and the long shot.
- Discuss each in one sentence: who would use it, and what is the riskiest thing about it?
- Pick one idea to test. You can keep a second one in reserve.
- Produce: your chosen idea in one sentence, starting 'For [user] who [need]...'
Solo
Vote with three dots yourself, then ask one family member to vote too. Fill the four categories before choosing.
Group
Each team member votes privately first to avoid copying, then the team decides. Use the decision method from ct-07 if stuck.
Related lessons: Making Good Decisions
Break
Rest, have a snack and prepare questions in your head for the next activity.
- Rest for 15 minutes.
- Think of three people who might have the problem you chose.
Your idea is a set of guesses
Before building anything, list what must be true for the idea to work, then test the riskiest part first.
- Write your assumptions as clear sentences: 'Students on long bus rides want to revise', 'They would use flashcards on their phone', 'They would pay something for it'.
- Put the riskiest guess first. Usually it is whether people really have the problem, not whether you can build the product.
- MIT's entrepreneurship centre quotes Eric Ries, author of The Lean Startup, defining a minimum viable product as the version that collects 'the maximum amount of validated learning about customers with the least effort'.
- Ask about past behaviour, not future promises: 'What did you do last time?' beats 'Would you use this?'
- Decide before testing what result would prove you wrong.
Related lessons: Testing an idea cheaply: customer conversations and simple experiments
Customer conversations
Real conversations with people who have the problem. Your aim is to learn, not to sell.
- Write a short opener: 'I am doing a project on how students revise on the bus. I am not selling anything. Can I ask you a few questions?'
- Hold at least three conversations of 8 to 10 minutes, only with people you know or people introduced by a trusted adult, never strangers in public or online. Ask: 'How do you deal with this now?', 'What is the most annoying part?', 'Have you ever spent money or time on a fix?', 'Who else should I talk to?'
- Only describe your idea in the last 2 minutes, and notice what people do, not just what they say.
- Write down exact words people use. They become your marketing language on Day 4.
- Record: for each person, their current workaround, how painful it is (1 to 5) and one quote.
- Produce: notes from at least three conversations and a verdict on your riskiest assumption.
Solo
Talk to relatives, friends or classmates by phone or in person, with a parent or guardian aware. Aim for five conversations by tomorrow.
Group
Teams interview students from other teams (who answer honestly as themselves) or invited adults. A supervising adult stays in the room.
Related lessons: Testing an idea cheaply: customer conversations and simple experiments, Writing professional emails and messages
Design a no-money experiment
Plan one small experiment that tests real interest without anyone paying you.
- Choose one: an interest list (a simple form where people leave a first name to hear more), a manual version (do the service by hand for three people for free), or a 'fake door' poster that explains the offer and counts sign-ups.
- Write the hypothesis: 'At least 5 out of 20 people who see this will sign up.'
- Decide how you will measure it and for how long.
- Check the safety and privacy rules: collect only what you need, keep it private and delete it after the camp.
- No real money: if anyone wants to pay, note it as strong evidence but do not take payment. Real selling comes later, only with a guardian and within the law.
- Produce: a one-paragraph experiment plan and the draft form, poster or message.
Solo
Run the experiment overnight within your family's network or class group chat, with permission from the group admin and your parent or guardian.
Group
Teams swap plans with another team for a 5-minute check: is it honest, safe and measurable?
Related lessons: Testing an idea cheaply: customer conversations and simple experiments, Identity theft and protecting your personal data, How scammers think: the red flags behind every scam
Evidence or opinion?
Sort what you heard today into evidence and opinion.
- List three things people did (evidence) and three things they said they might do (opinion).
- Decide whether your idea survives, changes or should be dropped.
- Complete the journal prompts.
Solo
Write for 10 minutes.
Group
Write for 6 minutes, then each team says in one sentence whether it is keeping, changing or dropping its idea.
Day 2 quiz
Five quick questions on ideation and testing.
- Answer the questions below.
- Read the explanations.
Block timer
Your journal
Private, saved on this device only. Your answers stay in this browser. They are never sent to us, even when you are signed in.
Day 2 quiz
Score 80% or more to complete Day 2.
1. During a d.school-style brainstorm, what should happen to judgement of ideas?
2. What target does the Bootleg suggest for a brainstorm?
3. Which is one of the d.school's four selection categories?
4. Whose definition of a minimum viable product does MIT Orbit quote?
5. Which result from an experiment is the strongest sign of real interest?
Overnight challenge
Run your experiment or hold two more customer conversations. Count the results honestly and bring them tomorrow, even if they are disappointing.