Interview stakeholders, properly.
Also called: Customer discovery interviews · Stakeholder conversations · Evidence interviews · Voice-of-customer research
Sitting down with the people who matter and asking honest questions, then listening for what they actually do rather than what they say.
Talk to the people whose verdict the product depends on: users, buyers, suppliers, experts. Ask about past behaviour, not future intentions. Avoid questions that fish for the answer you want. Stories about what someone actually did are evidence. Polite encouragement is not.
What interviewing stakeholders is
Think of it the way a good doctor takes a history. They do not walk in and announce “you have a chest infection, agree with me.” They ask what happened, when it started, what you tried, what made it worse. The diagnosis comes out of the answers, not out of the question. A stakeholder interview works the same way. You are not there to confirm your idea. You are there to collect a history, and let the diagnosis fall out of it.
A stakeholder is anyone whose verdict the product depends on. For a physical product that is wider than just the buyer. It is the people who would use it, the people who would stock it, the people who would make it, and the people who know the field well enough to spot a flaw you cannot see yet. Each one holds a different piece of the picture, and each one can sink the product on their own.
The single hardest discipline in the room is this: people are kind, and people are bad at predicting their own future behaviour. Ask “would you buy this?” and a warm person says yes to be nice. Ask “tell me about the last time this problem cost you something” and you get a real story with real detail. The first is a wish. The second is evidence. In my experience the gap between the two is where most products quietly die.
Who counts as a stakeholder
- Users. The people who would live with the product day to day. They tell you whether the problem is real and whether your fix actually fixes it.
- Buyers and retailers. Sometimes the same as the user, often not. A buyer at a chain like John Lewis or Lakeland decides whether a product ever reaches a shelf, and judges it on margin, returns and shelf space, not on how clever it is.
- Suppliers. The people who would make it. They tell you what is cheap, what is hard, and what will quietly wreck your costing six months from now.
- Experts. People who know the field. They save you from the mistakes everyone in that field already knows about and you do not.
You do not need a hundred interviews. You need the right handful from each group, asked well. Here is the interview plan we worked through for the proofing box, so you can see the shape of a good one rather than a generic template.
The two columns that matter most are the question columns. The difference between a useful interview and a flattering one comes down to how you ask. Set the two side by side and the trap is obvious.
Questions that gather evidence, and ones that don’t
A leading question hands the person the answer and asks them to nod. An evidence question asks them to remember something that actually happened. The first feels productive in the room and teaches you nothing. The second is often quieter and occasionally disappointing, which is exactly why it is worth trusting.
- “Wouldn’t it be great if your prove just worked overnight?”
- “Would you pay £149 for a box that solves this?”
- “Don’t you hate it when a loaf fails?”
- “Is a premium ceramic finish important to you?”
- “Tell me about the last time a prove didn’t work.”
- “What have you spent money on to fix this so far?”
- “How often does it happen in winter?”
- “Walk me through what your kitchen does overnight.”
Every question on the left can be answered “yes” by someone who will never buy. Every question on the right forces a real memory, and a real memory either backs your idea or quietly warns you off it. Both outcomes are worth more than a room full of polite agreement.
How it fits the bigger picture
Interview stakeholders is activity 02.10.06 in the framework, near the end of Stage 02 Discover. It feeds directly into the next activity, run idea brainstorm (02.10.07), where the evidence you have gathered becomes the raw material for generating and shaping options.
What it can do
It replaces opinion with evidence. A handful of well-run conversations tells you whether the problem is real, whether your fix lands, and whether the people who make and sell the product will play along. It surfaces the objection a buyer will raise long before that buyer can quietly decline to stock you.
What it can’t do
It can’t tell you what people will buy, only what they have done and felt. Stated intent is unreliable; a real purchase later is the only proof. And a few interviews are a signal, not a market. They point you in a direction; Stage 04 Evaluate is where that direction gets tested against numbers.
See the full 10-stage process →
Try it yourself
List the four stakeholder groups for your product: users, buyers, suppliers, experts. Find two real people in each. Write five questions, then cross out every one that could be answered with a polite “yes”. Replace each with “tell me about the last time…”. Record the conversations if you can, and write up what people did, separately from what they said they might do.
Or run the guided version. The Free Sprint walks you through who to talk to and what to ask, as a starter version of this activity. Start the Free Sprint →
Your interview checklist
Project notes: the conversations that changed the box
▸ From the notebook · optional reading
Dan and Anna Hartley spent two weeks talking to bakers, a ceramics maker in Stoke-on-Trent and a buyer. One throwaway answer reshaped the whole product.
3 min read · click to open
Dan wanted to send a survey. “Quick poll in the Sourdough School group, hundred replies by Friday, done.” We pushed back: a survey would give him a hundred polite yeses and nothing he could trust. Ten real conversations would teach him more. He grumbled, then booked the calls.
The bakers
The first instinct was to ask “would you buy a heated box?”. I asked him instead to open every call with: “Tell me about the last loaf that didn’t come out.” The room changed. One baker in a cold terraced kitchen described binning dough three weekends running through January, then buying a seedling heat mat off the internet and balancing a cardboard box on top of it. Another kept her oven light on all night and worried about the electricity. Nobody had been waiting for a clever gadget. They had all already spent money, badly, trying to solve it themselves. That was the evidence. The pain was real, recurring and already costing them.
The ceramics supplier
The conversation in Stoke-on-Trent was the one that nearly broke the costing. Dan went in describing the shape he wanted. The maker listened, picked the drawing up, turned it over and said the wall section was too thick to fire reliably at the price Dan had in his head, and the lid would warp. Half an hour with someone who actually fires ceramic saved months of arguing with a part that was never going to work. We redrew the shell that week.
The buyer
The kitchenware buyer was blunt in the most useful way. She had no interest in the temperature accuracy Dan was proud of. What she asked about was the return rate, the box on the shelf, and whether it needed an app, because “anything with an app and a winter, we get back in January”. The no-app decision, which Dan had half made on instinct, suddenly had a hard commercial reason behind it. I wrote her exact words at the top of the positioning notes.
Three groups, fewer than a dozen conversations, two weeks. The bakers proved the problem. The supplier saved the costing. The buyer handed us the positioning. None of it would have come out of a survey, and Dan never tried to send one again.
— Discover stage, project notes, 2026
— Next in Discover → Run idea brainstorm