Most family history projects do not fail during the work. They fail during the planning, in the weeks where somebody researches recording equipment, reads about archival storage, opens a spreadsheet of every relative they can name, and quietly stops.

The scope is what kills it. “Record the family history” is not a project, it is a category. You cannot start it, which means you cannot finish it, which means the whole thing sits at the same size forever and gets heavier every month.

What follows is the version that actually gets done.

Start with one person, and one hour

Pick a single relative. Not the oldest, not the one with the most dramatic life. Pick the one you find easiest to talk to, because your first attempt is going to be clumsy and you want it to be clumsy with somebody forgiving.

Then give yourself one hour with them. Not a project. One hour.

This does two things that a bigger plan cannot. It produces something real within a week, which is the only thing that makes a second session happen. And it teaches you, quickly and cheaply, what you are actually bad at: talking too much, jumping between decades, forgetting to ask what happened next.

Everything you plan before the first hour is a guess. Everything you plan after it is informed.

Decide what finished looks like

Write one sentence describing what you want to be holding at the end. Then stop planning.

The sentence matters because the three common versions of this project are completely different pieces of work:

A recording. Their voice, kept. Two or three hours of audio, backed up, labelled. The most faithful and the least likely to ever be listened to again.
A written account. Their life in their own words on paper, or in yours. Slower to make, far more likely to be read by somebody who never met them.
A documented tree. Names, dates, places, sources. This is genealogy, and it answers who and when rather than what it was like.

Most people want the second and start the third, because the third has obvious first steps and the second does not. If you are unsure which you want, read family tree or family stories, and which to do first.

The first hour, in detail

Sit somewhere they already sit. Their kitchen table beats your living room, and both beat anywhere that feels like an appointment.

Ask about one decade, not a life. “Tell me about when you were first married” is answerable. “Tell me about your life” is a request for a summary, and nobody can summarise sixty years on demand.

Bring five questions, expect to use two. The rest of the hour comes from following what they actually said, which is a skill worth reading up on properly in how to interview a family member.

Record it if you can do so without making it a production. A phone on a folded tea towel is enough, and the audio side is worth ten minutes of preparation, because a recording you cannot hear is not a recording.

Pick a format you will keep up, not the best one

The best format for preserving a life is a professionally produced oral history with a transcript, an index and archival storage. Almost nobody completes one.

The format that gets completed is the one that survives a bad week. Ask honestly: if you skip three weeks in November, does this project resume, or does it become a folder you avoid?

Recording is easy to start and hard to finish, because the work sits at the end, in the transcribing and the editing. Writing is hard to start and easy to finish, because each session produces a finished piece of the thing. A guided journal with printed questions sits in the middle and moves the effort onto the person whose story it is, which is either the whole point or the reason it stalls, depending on the relative.

What to do about everyone else

Nothing, yet. This is the part people get wrong.

Announcing a family history project to the whole family produces opinions, corrections, one relative who wants to take over, and one who is upset about the order you asked people in. Do the first hour privately. Show the result. The project becomes real to other people at the moment they can see one, and not before.

There is one exception. If somebody in the family already has a box of photographs or a half-finished tree, go and see it early, because you will otherwise spend six months rebuilding it.

An order that works

One hour with one person. This week.
Back it up the same day. Two places. This is where most projects lose their first session.
A second hour with the same person, within a fortnight, before the first has gone cold.
Write up one story properly. Not all of it. One story, start to finish, in a form somebody else could read.
Show it to one relative. This is the step that recruits help and settles the argument about whether it is worth doing.
Then widen it, to a second person or a longer arc.

Steps one to five take a month of ordinary evenings. Step six is where the project either has momentum or does not, and momentum is a separate problem worth solving deliberately: how to keep a family history project going.

Two things that will slow you down

Buying equipment. A modern phone records better audio than most consumer recorders sold a decade ago. If you eventually want better sound, you will know exactly what is wrong with what you have, which is a much better basis for buying anything.

Reading about method for longer than you spend doing it. There is a large, genuinely interesting literature on oral history. Read it after your third session, when it will make sense, rather than before your first, when it will mostly make you feel unqualified.

Start here

Text one relative and ask if you can come round for an hour on a specific day. Name the day. An open invitation to talk sometime is the same as no invitation.

Then write down five questions about one decade of their life, and let the other fifty-five wait. They are not going anywhere, and neither is the project, once it has actually begun.