Geotagging Photos for Genealogy and Family History
A family tree is also a map. Here is how to add location to genealogy photos, treat AI suggestions as research leads rather than proof, and get the coordinates into FamilySearch, Ancestry, MyHeritage, and other family-history tools.
A family tree is usually drawn as names and dates. But it is also a map. Every person on it was born somewhere, lived somewhere, and — often — moved. A surname clusters in a valley for two centuries and then scatters across an ocean in a single generation. Genealogy is, underneath the charts, a geographic discipline. And photos are the visual layer of that geography: the house, the street, the church, the harbor. This page is about getting those photos onto the map — adding location to family-history photos in a way that holds up to a researcher's standards.
TL;DR: Genealogy photos are worth geotagging because place is a first-class genealogical fact — records are organized by location, and migration is half the story. AI visual recognition can suggest where an undated, untagged photo was taken, but treat that suggestion as a research lead, not a citation: corroborate it against census records, registers, and family knowledge before you trust it. Once a location is confirmed, RetroTagr writes standard EXIF GPS into the file, and every major family-history tool — FamilySearch, Ancestry, MyHeritage, RootsMagic, Gramps — reads it automatically.
Why genealogy is a geographic discipline
Open any genealogical record set and the first thing it asks for is a place. Census records are enumerated by district. Parish registers belong to a specific church in a specific town. Immigration manifests are organized by port of departure and port of arrival. Land records, draft registrations, electoral rolls — all indexed by where, not just who. When you research an ancestor, you are constantly asking which town, which county, which parish, because that is how the evidence is filed.
Migration is the other half. A family that appears in one county's records for generations and then vanishes did not vanish — it moved, and the research problem becomes finding where. Tracing that movement is one of the core skills of genealogy.
Photographs sit naturally inside this geographic frame. A photo is evidence that a particular person was in a particular place at a particular time. Pinned to a map, a collection of family photos becomes a visual record of where a family actually lived its life — which is exactly the question a family tree is trying to answer. The geotag turns a loose image into a placed one.
Location is a research lead, not a citation
Here is the part that matters most for a genealogist, and the part where AI tools are most often oversold.
AI visual recognition can look at an old photo and suggest where it was taken. It reads architecture, signage, terrain, vehicles, and clothing era, and it produces a coordinate with a confidence band. That is genuinely useful — but a suggestion is not a source. It is a lead, in exactly the sense genealogy already uses that word: a hint that tells you where to look, generated without a citation.
The discipline is the same as for any unsourced hint. The AI says "this looks like a row house in a specific part of Boston around 1910." You do not write that into the tree as fact. You treat it as a hypothesis and you go to the records: was this branch of the family in Boston in 1910? Does the census put them at an address in that neighborhood? Does a sibling's birth record name the same parish? If the records agree, you now have a sourced location and the AI suggestion did its job — it pointed you at the right haystack. If the records disagree, you have learned something too.
Used this way, AI geotagging is a force multiplier for research, not a shortcut around it. It generates plausible leads across a whole batch of photos far faster than you could eyeball them, and you apply your normal evidentiary standard to each one.
The workflow for a researcher
Genealogy photos rarely arrive all at once. They come in waves — a cousin shares a scanned album, an archive fulfills a request, a parent finally hands over the box. The workflow is built around tagging each wave as a batch:
- Gather the batch. Whatever just arrived — a shared album, a folder of archive scans, a stack of phone photos of prints.
- Let the AI suggest. Upload the batch; the AI returns a location and a confidence band for each photo, or flags the ones it cannot read.
- Corroborate. This is the research step. For each suggestion, ask whether it fits what you know and what the records say. Accept the ones that check out, correct the ones that are close, flag the ones worth a record search, and skip the ones the AI could not place.
- Write the location. RetroTagr writes standard EXIF GPS tags into the photo files. Original timestamps and other metadata are left untouched — only the GPS tags are added.
- Bring them into the tree. Import the tagged photos into your genealogy software. The locations are already in the files, so they appear without extra steps.
The photo is now a placed piece of evidence, cross-referable against the place-organized records that genealogy runs on.
How geotagged photos flow into genealogy software
The reason EXIF GPS is the right format is interoperability. There is no genealogy-specific location standard for photos and there does not need to be — every major family-history tool already reads the GPS tags that cameras and phones have written for two decades.
FamilySearch Memories, Ancestry's photo features, MyHeritage, RootsMagic, Family Tree Maker, and Gramps all read standard JPEG EXIF. A scan that RetroTagr has tagged carries the same GPSLatitude and GPSLongitude fields a modern smartphone writes, so the genealogy tool cannot tell the difference between a tagged 1950s scan and a photo taken yesterday. You attach the photo to a person or an event as you normally would, and the location travels with it.
That interoperability is also insurance. Genealogists change tools over the years and pass research to the next generation. A location written into the photo file itself is not locked to one application — it stays with the image no matter which software opens it next.
What AI cannot place
Be clear-eyed about one thing: a large share of an old family archive is studio portraits and interior shots. A formal portrait against a painted backdrop, a christening photo in a front room, a tight headshot — these give a vision model nothing. There is no street, no skyline, no sign. The AI will flag them low-confidence or return nothing, and that is the correct behavior.
For those photos, the location comes from you. You may not know the exact address, but genealogy research usually tells you the town a person lived in during a given decade — and a town-level pin is still meaningful on a family map. Set it manually from what your records already establish.
The point of geotagging a family archive is not a perfect coordinate on every image. It is getting enough of the collection placed that the map becomes a real answer to where a family lived — a layer of evidence as legible as the dates on the tree.
If you want to try the workflow, RetroTagr's free tier covers 100 photos and 5 AI suggestions — enough to tag one shared album and see whether it earns a place in your research routine. For the digitization side of the same project — scanning prints into files in the first place — see the guide to geotagging scanned family photos.