Interview transcript template: how to find Grandma's voice again

Grandma sits in the living room on a Tuesday and talks about the move in 1968. The recording holds the laugh. The words beside it are only the map that gets you back to minute twelve. If you need an interview transcript template, you need three things on the page: who spoke, what they said in readable English, and a time you can jump to in the recording.

What the template should do, and what it should not

A family interview transcript template is not a research protocol. It should help a grandchild find the sentence about the move without listening to the whole hour again.

How you hold the conversation and keep the original file is in how to interview parents. This page is only the sheet afterward: how the writing is laid out.

Universities write long rulebooks for this. The Frankfurt University transcription guide splits a transcript into three parts: a header with the basic facts, the text itself, and often an appendix. At the kitchen table, the first two are enough. Skip the seating chart and the class period.

At the top of the first page: who, when, where

Before the first sentence, put a short header. Without it, a year from now you will not know which conversation this was.

This is enough for the family:

Conversation: Grandma talks about the move
Date: March 3, 2026
Place: living room
Who speaks: Anna, Grandma

You do not need more in the header. Anyone who reads it later should see who sat at the table. Labels like I and B are for studies. At home, the names stay.

Each person gets their own line

When Grandma answers, a new line starts. Leave a blank line between the two parts. That is what Kuckartz and Rädiker describe: one turn, one line, space in between so you see the change.

Put the name in bold, then the sentence. Write it so someone can read it aloud, not letter by letter from the sound. If Grandma says the words in a heavy dialect, write the meaning in plain English. You still hear the sound in the recording.

This is what the excerpt looks like:

Anna: What was it like back then, when you moved? [00:04:12]

Grandma: It was hard. We barely had any money. But my mother always said it would work out. [laughs] [00:04:24]

You may put a laugh or a long pause in brackets if it belongs to the story. You do not have to count every breath.

Research software often wants columns, line numbers, and hidden names. The ATLAS.ti guide to formatting and anonymizing interview transcripts explains that for studies that will be published. In the kitchen drawer, the real names stay. Grandma is Grandma, not B1.

The time mark, so you can jump in the recording

Without a time mark, the text is a nice write-up, but not a map. Put the time at the end of the stretch where someone finished speaking. Seconds are enough. [00:12:04] is enough to find the move again.

University projects often put a time mark before every speaker change and cut the text into short pieces, because those pieces become captions. The CeDiS guide at Freie Universität Berlin does that for video archives. You do not need that for a family conversation. One mark when the person changes is enough.

If you cannot hear a stretch, leave a gap. Do not write a guess as if it had been said. Listen again. The sentence in the recording is the one that counts.

The first text from an app is a draft

An automatic transcript often already has the names and the rough order. Stress, laughter, and mumbled words are usually missing, as Sonix explains in its formatting tips. So you do not paste the raw file as-is. You move it into your template: check the names, set the times, make the sentence about the move readable.

In FamilyStories that text is made on the device, not at an outside service. The original voice can still stay. The words beside it are the map. Questions that start the conversation are in questions to ask parents about their life.

How you fill the template at the kitchen table

Do it in this order. Then everything stays in the right place later.

1. Put the header in. Write the conversation, the date, the place, and the names. Otherwise you will open the wrong file in a year.
2. Listen for a minute. Are the voices clear enough? If not, sit closer before the whole talk sounds like that.
3. Write the text in the template. Each new voice gets a new line with the name, in readable English.
4. Set the times. Whenever someone else speaks.
5. Read it out loud once. Does a sentence still sound like Grandma? If the text feels stiff, listen to that spot again and make it spoken, without twisting the meaning.
6. Save it twice. One file for reading, such as Word. You save the recording on its own. The text does not replace the voice.

If you want to take the collection with you later, FamilyStories lets you do that on a paid plan as a Word or PDF file. The app does not make a printed book.

Better a transcript you can read at the kitchen table than one nobody can find again.

The text belongs on the same question as the voice

A neatly laid-out transcript in a folder with no name is, a year later, just another file. The story lives in the voice that tells the move.

FamilyStories does not follow university rules for transcripts. The app records the answer as speech, turns it into text on the device, and can keep the original voice beside it. FamilyStories cannot read the contents and cannot hear them, because they are end-to-end encrypted. The app runs on iPhone and iPad. The guest link is a key: whoever has it can read and edit open questions. Up to ten people can collect the family's shared memories in a private circle. Everyone can contribute: a name, a photo, a voice. The circle is there to collect, not to help.

Without FamilyStories, this guide is still a template for a header, names, and a time mark. What is then missing is the place where the text does not sit as a loose file, but on the same question as the voice.