Cohesion Lab
Menu

Free guide

A starter checklist for a family archive.

Seven things to do before you buy software, hire a service, or pay for storage. Most of this work survives the choices you make later.

Why this guide exists

Family archives are made out of objects with stories: photographs, recipes, letters, audio reels, video tapes, oral histories, ephemera. Most of them have at most one person who knows what they are. The first job of stewardship is making sure that knowledge doesn't disappear with that person.

This is the work you can do — and should do — before deciding what software, service, or platform to use. None of it is locked to a vendor. All of it makes the next step easier.

Step 1

Inventory before software.

Before any tool, list what you have. A spreadsheet is fine. Three columns at minimum: type (photo, letter, recipe, audio, object), rough count, and condition (fragile, stable, already digitized).

Mark the most fragile materials separately — paper before 1980, magnetic tape, film. These are the things that need acid-free storage and digitization first; everything else can wait.

Flag undated and unidentified items. They are the ones to ask the oldest people in the family about — soon, while you still can.

Step 2

Name things consistently.

Pick a naming convention for files and stick to it. A defensible default: YYYY-MM-DD_subject_creator_version.ext (e.g. 1962-08_summer-cabin_pearl-henderson_master.tif).

Pick one form of every name and stick with that, too. Aunt Pearl and Pearl May Henderson are the same person; the archive needs to know that. A "names" tab in your spreadsheet — known-as, full name, dates, role in the family — is worth more than any software feature.

Use a small controlled vocabulary for relationships. "Daughter of X" is searchable; "X's kid" is not.

Step 3

Provenance is a story. Write it down.

For every significant item, write one to three sentences: what is it, where did it come from, how do we know.

Be honest about gaps. "We don't know who took this photo; it was found in Grandmother's box when we cleared the attic in 1998" is a real piece of provenance. Pretending you know is worse than admitting you don't.

Date the note itself. Your understanding of an item will change over time, and a future you needs to know which version of the story this paragraph is.

Step 4

Decide who sees what.

Family archives have layers of audience. Sketch them now: public (open to the world), family (anyone who can prove descent), immediate family (those alive today and their children), just you (sensitive material — medical histories, custody disputes, private letters).

Tag every item with one of those tiers. The tags are the seed of any future access system, and they make the conversations you'll eventually have with relatives — about putting some of this online — much easier.

Make sure at least one other person knows the access decisions. A locked archive nobody else can open is the same as no archive.

Step 5

Pick formats that outlive software.

For each material type, pick an open archival format:

  • Images: TIFF for archival masters; JPEG for sharing copies.
  • Documents: PDF/A for scanned originals; plain text or Markdown for transcribed text.
  • Audio: WAV (uncompressed) for archival; FLAC or MP3 for sharing.
  • Video: FFV1 or uncompressed MOV/MKV for archival; MP4/H.264 for sharing.
  • Spreadsheets: CSV alongside any .xlsx — CSV will outlive Excel. (CSV preserves the values in a spreadsheet, not its formulas or formatting.)

Avoid formats locked to a single application (.psd, .ai, .pages, older proprietary .doc variants). They can't be relied on to stay readable over time, and your archive will outlive any single application.

Step 6

Back up the masters.

The archival rule is 3-2-1: three copies, on two different kinds of media, with one of them off-site. For a family archive that means, for example: an external hard drive, a cloud backup, and a second hard drive at a sibling's house. 3-2-1 protects against loss, but not silent corruption — so also keep a checksum (a fingerprint) of your master files and re-check it periodically.

Test recovery once a year. Pick a random file and try to open it from the off-site copy. If you can't, the backup is theoretical — which is the same as no backup.

Step 7

Write a README.

One plain-text or Markdown file at the top of the archive, named README.md. Keep it short. Cover four things:

  • What's in here. Two or three paragraphs.
  • How it's organized. Folder structure, naming convention.
  • Access tiers. Who is allowed to see what.
  • Who to contact. If something happens to you, who picks this up.

The README is the document a future caretaker reads first. Write it for them, not for yourself.

That's the work that doesn't depend on us.

If you finish this checklist and decide you'd like help going further — cataloging at standards, building a display, sharing selected material with the family or the public — that's our work. If you finish this checklist and don't need us, that's fine too. Either way, the archive is in better shape than it was.