HEINZ BRO’S

Heinz Bros

Private family archives · described, sourced and kept

Standards

The standards this field publishes, and where we go past them.

Genealogy has written standards. Almost nothing sold to families about their own photographs observes them. Below is each standard we hold ourselves to, what it actually requires, and what we do about it.

Kept means we meet the standard as it is written. Exceeded means we do something it does not ask for — not that the standard is wanting.

The Genealogical Proof Standard

Kept

Requires: reasonably exhaustive research; complete and accurate citations; thorough analysis and correlation; resolution of conflicting evidence; and a soundly reasoned, coherently written conclusion.

Every conclusion we hand you has been through all five. Where a document corrects itself — an amended age, a struck-through year — both readings are recorded and the tie is broken from outside that document, because arithmetic that works is not the same as arithmetic that is right.

Genealogy Standards, Board for Certification of Genealogists

Kept

Requires: planned research, documented as it is done; sources recorded at the moment of use; and conclusions revised in the open when new evidence arrives.

Research is scoped in writing before it starts. Findings are staged as they are read and written up in one pass, so nothing survives only in somebody’s memory of a conversation. Nothing is ever deleted — a withdrawn conclusion is reclassified and kept, with the reason.

Evidence Explained, Elizabeth Shown Mills

Kept

Requires: citations assembled in a fixed element order, and a clear separation between the source, the information in it, and the evidence it provides.

Citations are built in her element order — creator, jurisdiction, record, then how it reached us. And where one document carries facts from different mouths, we name them: on a death certificate the date of death is the physician’s, while the birth date and the parents are a grieving relative’s. One page, two informants, two weights.

Two marks on every record, never one

Exceeded

Required by nothing. Common practice is a single confidence label, if any.

We keep how sure we are apart from what we actually examined, and neither may prop up the other. An index entry can never be called Documented, however plausible it is. A copy of a copy promotes nothing. An inscription in a hand nobody can identify settles no face.

A search that found nothing is a deliverable

Exceeded

Required by nothing. An empty search is usually invisible, and so it is repeated.

Every record class searched without result is written up: what was looked at, over what span, and why it was reasonable to expect something. You paid for that search once. Nobody — including us, two years later — should pay for it again.

Provenance and original order

Kept

Archival practice requires: records kept in the grouping and order they arrived in, described rather than rearranged, with their custody recorded.

Photographs are filed by where they came from — the envelope, the album, the relative, the date it passed to you — and never by who is in them. Browsing by person or by decade is a view generated on top of that, never a folder. Where an album was broken up before we ever saw it, that order is simply gone, and we say so rather than invent one: two photographs sitting together in a box is not evidence that the people in them are related.

Fixity and digital preservation

Kept

The NDSA Levels of Digital Preservation require: more than one copy in more than one place; checksums recorded and re-verified; open formats; and metadata that travels with the file.

Every image is checksummed as it arrives. A byte-identical duplicate is refused; a second print of the same photograph is accessioned in full and linked to the first, because two versos in two hands are two witnesses. Copies live in more than one place, the store is re-verified on a schedule, and each image’s description is a plain-text file beside it — readable in fifty years by anything that can read a text file.

The professional code for client work

Kept

The Association of Professional Genealogists’ code of ethics requires: a written agreement before work begins, no guarantee of a particular result, and client confidentiality.

No membership obliges us to follow it. We follow it. Scope in writing first; the possibility of finding nothing stated before you engage us, not after; and no client named, shown or quoted without their specific written permission.

The privacy of living people

Exceeded

Required by nothing, for a private family site.

Living people are withheld by default and are absent from a published site rather than blacked out in it — a redaction still announces that somebody is there. The data behind the site is built the same way: what is withheld is recorded, the value itself is kept somewhere the site cannot reach, and the site is correct with or without it.

The two marks

How sure, and from what

Every identification and every conclusion carries both of these, and they are set independently.

  • How strong the conclusion is. Documented · Probable · Unproven · Eliminated. Recorded for each person in a photograph separately, because one face in a group can be certain while the eleven beside it are guesses.
  • What was actually examined. The original print · a digital original · a scan only · a copy of a copy. For records: the image itself, an index entry, a derivative, or an older compilation of unknown basis.

An eliminated candidate stays on the record for good. Three men of the same unusual surname were ruled out of one of our own lines — one of them sharing a forename, a middle initial, a brother’s name and a birth date inside the right window. They are written down precisely so nobody rediscovers them in ten years as good news.

Worked example

What all of that looks like on one photograph

This is an example — not a client’s — of the record that comes back to you with a single print, and that travels with it for the rest of its life.

Photograph record

IMG0147 · accessioned 14 Sept 2026
Provenance
Brown paper envelope, hall closet; from the estate of the client’s grandmother. Container photographed and described at intake.
Subjects
Standing, left — Albert R. · named on the verso in his sister’s hand, and the hand is identifiedDocumented
Seated, centre — his mother · identified by the client from family knowledge; told to her by an aunt, 1994Probable
Standing, right — not identified · a family resemblance and nothing furtherUnproven
Date
Between 1911 and 1918 — from the photographer’s imprint, a studio in business those years, with the sitters’ apparent ages.
Place
Written on the mount: “Quincy.” Inferred: Quincy, Adams County, Illinois — untested. A photographer’s imprint is a street address, not a date and not a county.
Source state
Original print seen.
Fixity
sha256 3f9c…a17b — re-verified on every backup, because silent corruption is the failure nobody notices.
Withheld
One living person named on the verso. Absent from the published site, not blacked out in it.

Nothing in this record is a guess presented as a fact, and nothing true about it has been left out because it was inconvenient. That is the whole standard, on one photograph.

House rules

Five rules that do the actual work

  • Provenance is captured before anything else. Which envelope, which closet, which relative, and when it passed to you. That is the half that dies with memory, and no amount of scanning brings it back.
  • A place-name is not a jurisdiction. What the document says, then what we infer from it, then whether the inference has been tested — kept apart in that order, so no later conclusion quietly inherits the middle step as fact.
  • An age is only as good as its anchor. A census age is the age on that census’s own reference date, not the day the enumerator called. Every record in a comparison is computed the same way, and we say which way.
  • A rare surname is not an identifier. Before a match is accepted, we name the fact that would disprove it and go and check that fact. “There cannot be two of him” is a feeling, not a finding.
  • Nothing is deleted, only reclassified. Withdrawn conclusions, eliminated candidates and corrections all stay on the record, with the reason and the date.

Plainly

On certification

Neither of us currently holds a credential from the Board for Certification of Genealogists or from ICAPGen. We are not going to imply otherwise on a website.

A credential is a periodic examination of a person’s work by their peers, and it is worth something. What we offer in its place is narrower and more direct: our own work is open to you before you hire us — the compilations, the revision histories that show every correction we have ever made, and a register of two hundred numbered notes on the ways this work goes wrong, most of them written after it went wrong on us.

And if your matter is one where a credential is genuinely required — a lineage society application, a citizenship claim, evidence for a court — we will tell you so at the first conversation and help you find someone who holds one.

Ask us a hard question about any of this

How to reach us →