These are two different activities that share a name, and confusing them is the most common way a family history project ends up producing nothing anybody wants to read.

A family tree answers who and when. Names, dates, places, and the documents that prove them. It is research. The material is in archives, registries and census records, and most of it will still be there in twenty years.

Family stories answer what it was like. What the house smelled like, why he left, what she was afraid of, how they actually met rather than the version told at the wedding. The material is inside living people and it is not held anywhere else.

That difference sets the order, and the order is not a matter of taste.

Do the stories first

Not because they are more valuable. Because they have a deadline and the tree does not.

A parish register from 1841 will be readable in 2050. Your grandmother’s account of the shop on Mill Street exists in exactly one place, and it is not a place with a preservation policy.

The tree is patient and people are not. That is the entire argument, and it holds even if genealogy is the part you actually enjoy.

There is also a practical reason to go this way round. The person who can identify everyone in the photographs, settle which of two women is which, and explain why the family moved in 1958 is a research resource that no archive holds, and they are considerably easier to consult than a parish register.

Why the tree tends to win anyway

Because it is much easier to start, and because it rewards you immediately.

Genealogy has obvious first steps. You type a name, something comes back, you have made progress within ten minutes, and there is always another record to find. It is genuinely absorbing and it produces a visible artefact that grows.

Interviewing a relative has none of that. You have to arrange it, you have to be in the room, it is socially awkward the first time, and the result is an hour of audio that looks like nothing until you do more work on it.

So people intend to do both, start with the tree because it is available on a Tuesday evening, and are still five generations back four years later with no recordings at all.

If you recognise that pattern, the fix is a deadline on the human part, not more discipline about the tree. Book one conversation, on a named day, this month.

What each one is actually good for

The tree gives you:

Names and dates you can verify.
The shape of the family, including the parts that were never mentioned.
Places, which unlock everything else. Knowing which village somebody came from turns a vague story into a specific one.
Corrections. Family memory is confidently wrong about dates with impressive regularity.
Occupations, which tell you a great deal about a life nobody described.

The stories give you:

Motive. A tree shows that somebody emigrated in 1912. Only a person can tell you why.
Texture. What things cost, what was eaten, what was frightening, what was funny.
Voice. The way a person spoke, which no record holds.
The things that were never written down anywhere, including most of what families actually want to know.
The parts documents actively hide. A record says “married”. It does not say it was arranged, or unhappy, or a second marriage nobody mentioned.

Neither is complete. A tree without stories is a list of strangers. Stories without a tree drift loose from time and place and become anecdotes.

Use each to feed the other

Once you have both running, they improve each other quickly.

Take documents to the interview. A census entry showing seven people in four rooms is a far better prompt than any question you could invent. Put it on the table and ask what that was like.

Take the stories to the archive. “She said they came over from Cork after her father died” gives you a name, a place and a date range to search, which is exactly what you need to find a record.

Use the tree to correct dates, not memories. When a relative says 1961 and the certificate says 1963, record both. The document is evidence of what happened. The memory is evidence of how it was held, and quietly overwriting the second with the first destroys something you cannot get back. This is worth doing visibly, in square brackets, as described in how to transcribe a family interview.

Use photographs as the bridge. They contain names for the tree and prompts for the stories, and they are the single most efficient thing to put in front of an older relative: how to label and preserve old family photos.

A workable order

This month: one conversation with the oldest relative you can reach. One hour. Recorded if possible.
Same month: the photograph box, with that person, in pencil.
Then the tree, starting from what you were told rather than from a blank search. Work backwards from yourself, one generation at a time, and prove each link before moving to the next.
Back to the people, with what the documents turned up. This second round is always better than the first, because you have specifics to ask about.
Repeat. The two halves alternate indefinitely.

If you are starting from nothing at all, the first-week version of this is in how to start a family history project.

One warning about the tree

It is very easy to accept somebody else’s research because it is there and it matches. Public trees are full of confident errors that have been copied hundreds of times, and a wrong link four generations back invalidates everything above it.

Prove each generation with a document before you build on it, and note where each fact came from. Being right and slow is better than being fast and building a tree of somebody else’s mistakes.

The same rule applies to the stories, incidentally. Write down who told you, and when. In fifty years that line is what lets a relative judge how much to trust the page.