What to Include in a Genealogy Research Log (The 12 Columns That Actually Earn Their Place)

Most genealogy research logs fail in the same way: they become a second, worse copy of the citation list.

You find something, you feel good, you write it down. You spend four hours in a database and find nothing, you feel bad, and you write down nothing at all. Eighteen months later you search that same database again, because there is no record anywhere that you already did.

That is the whole problem in one paragraph, and the fix is a single column.

Here is a log layout that survives contact with real research, and the reasoning behind each column that earns a place in it.

The Copy-Ready Layout

Twelve columns. Four of them do work that no citation list does.

Column 8: The One That Matters Most

Result should have NOTHING FOUND as a first-class value, sitting in the same column as your successes and appearing in the same list.

This is the column that separates a research log from a trophy cabinet.

A negative result is genuine, costly research output. Searching the 1868 register for a baptism that is not there took real hours and possibly real money, and it produced real information: the child was not baptised in that parish in that year, which narrows where they were. That finding is only worth something if it survives, and it only survives if it is written down as deliberately as a success.

Without it, your research has no memory. With it, your log starts answering the question that actually blocks progress — not what have I found? but what have I already ruled out?

Log it with the same care as a hit. Same date, same repository, same search terms, same next action.

Column 7: Search Terms, Including the Wrong Spellings

Record exactly what you typed, wildcards included.

Hale, Hale*, Hail, Heal — 1866–1872 is a different search from Hale — 1866–1872, and a “nothing found” against the first is worth far more than a “nothing found” against the second. Six months later, your own log entry has to be able to tell you how thoroughly you looked, or the negative result is not reusable.

This is also where the misspellings get preserved. Enumerators, clerks and priests wrote names by ear. The variant spellings are not noise to be cleaned up — they are the search terms that will eventually find the next record, and they deserve to be stored somewhere you can find them again.

Column 9: What It Means

One sentence, written in the moment, converting a raw result into a conclusion.

“Rules out St Finbarr; try neighbouring parishes” is a different thing from “nothing found”. The first is a decision made while the context was still in your head; the second is a fact you will have to re-reason from scratch. The sentence costs fifteen seconds to write and saves twenty minutes of reconstruction every time you come back to the line.

Column 10: Confidence

Four levels, attached to the record rather than the person:

• Proven — a primary record, contemporary to the event, that names the person unambiguously.
• Probable — strong evidence with a small leap in it. A census that fits every detail but does not name a parent.
• Possible — a plausible candidate you have not confirmed. The right name, the right decade, the wrong certainty.
• Disproven — actively ruled out.

Two things make this column work.

It attaches to the record, not the person. One ancestor can rest on one solid record and three weak ones, and averaging that into a single rating for the person hides exactly the information you need.

Disproven records stay in the log. Do not delete them. The point of keeping a disproven record is that in two years you will find that same promising lead again, feel a small thrill, and your own log will be there to say you already checked, and it was not him. Deleting it guarantees you will re-run the excitement and the work.

The reason this column exists at all is that genealogy’s characteristic failure is not a wrong record — it is a Possible that nobody ever downgraded, sitting in a tree for six years until repetition has quietly promoted it to a fact.

Columns 11 and 12: Next Action and Target Date

A log without a next action is a diary.

Every open line should name the specific thing to do — a named record set, a named archive, a specific letter to write — and carry a date you intend to do it by. Not because the date is binding, but because a date makes an overdue list possible, and an overdue list is the difference between research and browsing.

The genuinely useful version of this is a formula that counts days past the target date and surfaces the worst offenders first, so that sitting down to research starts with a ranked list rather than a vague feeling about which line was interesting last time.

What Does Not Belong in the Log

Three things people put in and shouldn’t:

The people themselves. Names, dates and parentage live on the individuals table, entered once, with everything else reading from it. A log that duplicates person data will eventually disagree with itself.

Long transcriptions. Transcribe into a notes field attached to the source record, not into the log line. The log is an index of activity, and a line that runs to four hundred words stops being scannable, which is its only job.

Speculation dressed as a result. “Probably the same family” is a confidence rating, not a finding. Put it in column 10 where it can be revisited, not in column 8 where it looks like evidence.

The Two Deadlines That Behave Differently

Worth understanding because it changes how you rank the list.

Records are patient. Parish registers, censuses and civil registration will still be there in ten years. A record search postponed is a delay, not a loss. Some record sets do open on schedules — the United States releases census returns 72 years after the count, for example — but the direction of travel is that more becomes available over time, not less.

People are not. The aunt who knows who is in the 1952 photograph, which brother emigrated first, and what the family actually called the great-grandmother whose name you cannot find — that is the only research window in genealogy that closes permanently, and it closes without notice.

This is why a good research file ranks living relatives by age separately from the record-search list, and raises anyone past an age threshold you set who has not been interviewed yet. It is the only genealogy deadline that cannot be extended, and it is the one that routinely loses to the more satisfying work of chasing a record.

If your log has one bias, make it this one.

Where the log sits in the rest of the file — the individuals table it reads from, the pedigree it fills in and the sources it feeds: how to build a family tree spreadsheet that still calculates at five generations.

Featured on ReadySheetGo

Family Tree, Genealogy Research & Ancestry Record Organizer — $13.99

A dedicated Research Log tab holding the open question, the record to search next, the archive to contact and a target date — and recording what you have already searched and found nothing in, so the same dead end is never paid for twice. Overdue tasks are surfaced on the Dashboard automatically.

Alongside it, a Sources tab rating every record Proven, Probable, Possible or Disproven, attached to a person ID, which flags a record with no citation text, a person ID that does not exist in the sheet, and anything still being relied on after you disproved it.

Then the rest of the file the log feeds: Individuals with a completeness score out of eight and six row checks; a five-generation Pedigree with 31 ahnentafel-numbered slots built from the Father ID and Mother ID columns; DNA Matches with a Next Action column; Census Grid deriving the birth year a recorded age implies and flagging the gap; Timeline with age-at-event checks; Interviews ranking living relatives by age and raising anyone past the urgency age you set; Heirlooms; Research Costs with renewal warnings; and a Dashboard running 24 checks against your own entries.

16 linked tabs, 15,700 working formulas, and a fictional sample family already loaded — 30 people, 26 sources, 23 census rows, a tree reaching 1836 — so you can see it running before you type anything.

Works in Excel and Google Sheets. No macros, no add-ons, no subscription, works offline.

Get the Family Tree & Genealogy Research Organizer →