Interviewing Your Grandma? Don't Start With Her Childhood (and Skip the 100-Question Lists)

Here’s how most grandparent interviews actually go. You sit down, phone recording, list of questions in hand. You start at the obvious beginning: “So… tell me about your childhood.” She thinks for a second and says, “Oh, it was normal. We were poor. What else do you want to know?”

Ten seconds, maybe fifteen. You’ve got a page of questions left and she’s already reaching for the TV remote.

Nobody warns you about this. The internet hands you lists of 99, 100, even 158 questions to ask your grandparents, as if the problem were running out of things to say. It isn’t. The problem is that a list is not an interview. A bigger list just means more ten-second answers.

This is the first of three guides on doing it right: this one is what to do before you sit down, the second is what to do when she goes quiet on you, and the third is how to turn an hour of audio into something you can actually hand in or keep. None of it requires special equipment or any money. It requires about twenty minutes of prep that almost nobody does.

One thing before we start, so you know where this comes from: I build an AI memoirist that has interviewed hundreds of elders over WhatsApp, in ten languages. Building it meant getting the same things wrong that you’re about to get wrong, hundreds of times, and redesigning the questions until they stopped failing. (What our users actually say stays private, always. The lessons about HOW to ask are ours to share, and that’s what this series is.)

How many questions should I actually ask?

Eight to ten. Not a hundred. A question list is a menu, and nobody eats the whole menu.

When you walk in with 50 questions, two things happen. First, you rush. Somewhere in your head there’s a completion bar, and every long answer starts to feel like a delay. You stop listening and start waiting for her to finish so you can ask the next one. She notices. Everyone notices.

Second, the questions scatter. One about school, one about her wedding, one about the war, one about her favorite food. Each jump resets her memory to zero. She never gets to sink into any one part of her life, so every answer stays at the surface: “it was nice,” “it was hard,” “I don’t really remember.”

The sessions that produce rich stories are almost never the ones that cover the most topics. They’re the ones that stay somewhere long enough for the details to come back. Memory works like a dark room: it takes a while for your eyes to adjust, and every topic change flips the lights back on. When we built our interviewer, the single biggest improvement wasn’t a better question. It was making it stay on the same memory for one more follow-up instead of moving on.

The lists are still useful, but as a quarry, not a script. FamilySearch’s is a good organized one, and even StoryWorth’s own list quarries well. Here’s how to pick from whichever you’ve got.

Which 8 questions do I pick from the lists?

Two sensory warm-ups, four or five on one single part of her life, one about a person she loved, one that looks forward. Take whatever list you found and choose:

• Two warm-ups that touch the senses. Not “what was your childhood like” but “what did your street smell like in summer?” or “what was on the table at dinner when you were eight?” You’ll see why in a moment.
• Four or five questions about ONE part of her life. Pick a single territory: her first job, or how she met your grandfather, or the year the family moved. One deep place beats five shallow ones, and you can always cover another territory next time. (There should be a next time. Also in a moment.)
• One question about a person she loved. Her mother is the classic. Ask any oral historian: people who have little to say about themselves become unstoppable about the people who raised them.
• One closing question that looks forward, not back. “What do you want your great-grandchildren to know about you?” It ends the session on meaning instead of fatigue, and she’ll walk away feeling like it mattered. Which it did.

That’s the whole selection method. Twenty minutes with any question list and a highlighter.

What order should I ask them in?

Not chronological order. This is where I most disagree with the standard advice. The most common tip on the internet is “go chronological”: birth, childhood, school, marriage, and so on. It sounds sensible and it fails constantly. Chronology front-loads the interview with the part of life she remembers in the least detail (early childhood) phrased in the most abstract way (“tell me about growing up”). You’re asking her to do the hardest memory work first, cold, while she’s still self-conscious about the phone recording her.

Concrete beats abstract for memory retrieval. Every archivist manual, from FamilySearch’s oral history guide on down, says some version of “start with easy, non-threatening questions,” and the professionals at the Boston Public Library’s oral history program put it plainly: save the hard stuff until you’ve both settled in. But “easy” doesn’t mean “chronologically first.” It means sensory and specific.

So open with the kitchen, the street, the smell of the bakery, the sound of her school bell. Watch what happens. A specific sense-memory pulls the whole scene up with it, and suddenly she’s not answering a question, she’s there, telling you who else was in the room. That’s the state you’re trying to create. One good sensory question does more than thirty biographical ones.

Then, once she’s inside a memory, move to your one deep territory and stay there. Your questions are a rough path, not rails. If she wanders from her first job into a story about a co-worker who cheated at cards, follow the co-worker. Everyone who has done this more than once says the same thing: the best material arrives sideways.

And keep the heavy questions (regrets, losses, the war) for the last third, once trust is warm, or for a later session entirely.

How long should the interview be? Is one sitting enough?

Thirty to forty-five minutes, and no, plan for three sittings. Here’s the thing nobody puts on the pretty printable lists: talking about your own life for two hours is exhausting. It’s emotional labor. Professional oral historians cap sessions at 60 to 90 minutes for exactly this reason, and for a grandparent in her eighties, 30 to 45 minutes is the sweet spot.

The families who capture the most stories don’t do marathons. They do three short sittings: same chair, same tea, a few days apart. One granddaughter who turned her interviews into a printed book settled on ten-to-twenty-minute calls with one anchor question each, and let curiosity do the rest. Between sessions her memory keeps working on its own. She will start session two with “you know, I was thinking about what you asked,” and that sentence is usually where the best story of the whole project comes from.

If this is for a school assignment, two sessions is the practical minimum: one to collect, a short second one to fill the holes you only notice when you re-listen. (The Oral History Association recommends student projects capture at least 20 to 30 minutes of usable material, which two short sittings clears comfortably.)

Should I show her the questions in advance?

Yes, but only three of them, the night before. Not the whole list. Send the whole list and she’ll rehearse; you’ll get the official version of her life, the one with the boring parts pre-approved. Send nothing and the first ten minutes go to warming up cold memory. Three questions the night before gives her brain time to wander the old streets without turning it into homework.

A voice note works better than a text for this, by the way. It sets the tone: this is a conversation, not a form.

How do I record it without making it weird?

Phone four feet away on something soft, permission asked on tape, then forget it exists. Keep this part boring and it will work:

• Phone about four feet away, on a folded towel or cushion (it kills the table vibration), screen down so neither of you watches the timer.
• Ask permission ON the recording: “Is it okay that I’m recording this for the family?” One sentence. It’s respectful, and one day that clip of her saying “of course, sweetheart” will be worth more than half the interview.
• Record 10 seconds, play it back, check you can hear her, delete, start for real.
• Then forget the phone exists. Your grandmother will, if you do.

Before you go set up the tea

The prep, complete: 8 to 10 questions picked with the recipe, sensory opener first, one deep territory, heavy stuff last or later, three questions sent the night before, phone on a towel. Twenty minutes of work, and you’ve already avoided the mistakes that flatten most family interviews before the first answer.

What the prep can’t do is handle the moment she looks at you and says, “I don’t remember. Nothing interesting ever happened to me anyway.” That sentence has ended a thousand interviews. It shouldn’t, because it’s almost never true, and there’s a reliable way through it. That’s Part 2.

One last thing. People plan these interviews for years and do them never. If the assignment is what finally gets you into that chair, the assignment might quietly be the most valuable homework you’ll ever get. Set the date this week.