Getting good input
Subtly different questions can unlock significantly different answers. If you're preparing for a conversation with residents, read this first, then adapt the sample interview guide for your own project.
What makes this hard
People are nice.
They tell you what they think you want to hear.
People don't know.
Not because they aren't thoughtful, but because it's hard to describe something that doesn't exist yet. Ask residents whether they'd use a new savings program and most will say sure. Far fewer will sign up. The reason why is what we're trying to figure out.
People say one thing and mean another.
What people tell you is shaped by how they want to be seen and by what they've gotten used to. Someone who has been working around a problem for years will often tell you it isn't a problem, because to them it no longer is.
Explaining things is hard.
People give a quick account that leaves out the nuance and details that matter.
Asking questions
Ask what they understood, not whether they like it.
People will tell you they like almost anything. It's more helpful to hear their thoughts when they see something so you can read between the lines.
"Do you like this flyer?"
"What is this flyer telling you? What would you do if you saw this?"
Ask about one instance, not general habits.
A general answer is a summary, and summaries leave out the detail you need.
"How do you usually handle unexpected expenses?"
"Walk me through the last time something unexpected cost you money."
Ask about the past, not the future.
People are poor predictors of their own behavior, and they'll be polite if you ask them whether they'll use something new. Instead of asking if they'd try something, ask about something that already happened.
"Would you use a savings program if we offered one?"
"Tell me about the last time you tried to set money aside. In what ways would a program like this have made that easier? In what ways would it not have helped?"
Ask what happened, not why they didn't.
Questions that presume a failure get you a defense or rationalization. Questions about events get you a story with real barriers and reasons. Assume the person already has a strategy and ask what it is.
"Why haven't you signed up?"
"Tell me what happened when you tried."
Additional strategies
Ask to see what they reference.
People's descriptions leave out key details, and seeing what they're referencing or how they do something can reveal key omissions and insights. Whenever someone describes something they do, ask if they can show you. Where do you keep that? Can I see the last one you got? Walk me through what you did.
Follow up more than feels natural.
The first answer is the surface. Ask why, then ask again. Three or four follow-ups on one answer can reveal much more than initially meets the eye.
Let silence sit.
People often fill a pause with the thing they were deciding whether to say.
Don't defend your work.
If someone misreads your flyer, that's the finding. Note it and move on rather than explaining what you meant.
Don't correct them.
If a resident has the eligibility rules wrong, you've learned something important. Clear it up at the end if you think it might be useful information for them, but not in the moment. We want people to speak freely without worrying about whether they're getting things wrong.
Invite stories.
A story is often the best source of data. It's something that actually happened to them and contains context and details that are difficult to elicit through questions alone. Whenever possible, ask if people can tell you about the last time x happened.
Use their words.
When they name something, adopt their term for the rest of the conversation. The vocabulary residents use is one of the most valuable things you'll take away, and using a different term than what they used can unintentionally come across as a correction.
Be on the lookout for opportunities to dig deeper.
Compensating behaviors and system hacks
How are people working around what doesn't work? Where do they keep their money instead of a bank account? Who do they go to when they have a question? How do they keep track of important documents?
Follow up: "How did you come up with that?"
Points of friction or tension
What feels awkward, difficult, or alienating?
Follow up: "What about that didn't sit right?"
Points of ease
What feels natural and easy?
Follow up: "Why do you think that works so well for you?"
Surprises and challenged assumptions
Moments where you didn't expect the answer.
Follow up: "Can you say more about that? I wouldn't have guessed that."
Contradictions
When values and behaviors don't line up. Don't challenge the person or make them feel defensive. Ask them to say more.
Follow up: "Help me understand how those two things go together."
Underlying motivation
Listen for the why beneath the action.
Follow up: "What were you hoping would happen?"
Emotional language
Listen for emotion-laden language, like "that really stuck with me" or "I don't like it when."
Follow up: "Can you tell me more about that moment?"
Practical notes
Location
Meet somewhere the person already goes, somewhere they picked, or neutral locations like libraries, community centers, or cafes. City buildings may feel too official.
Length
Forty-five minutes to an hour for a one-on-one interview, about 90 minutes for a group session. Feedback interviews on draft materials can be twenty to thirty minutes. Say the length up front and stick to it.
Consent and recording
Create a brief, easy-to-read consent form (example here) that explains key information. Ask about recording before you start, and explain what happens to the recording. If you're not recording, bring a second person to take notes so whoever is asking questions can actually listen.
Who comes
Try to keep it to two people from your side. Too many people on one side of the table stops feeling like a conversation and makes it harder for participants to open up.
Write it up the same day
A week later you'll forget many of the details that made the conversation useful.
Pay people
Compensate participants for their time. It's a research cost, and it changes who is able to say yes.