Records two people share

Two people, one business or one household. Both capture receipts, both mean well, and at the end of the year there are two overlapping partial archives with an unknown gap between them. Each person assumed the other had the thing that neither did.

Shared recordkeeping fails differently from solo recordkeeping. The mechanism is not laziness — it’s that responsibility for any given item is ambiguous, and ambiguity resolves to nobody.

The four failures

Diffusion. A receipt in a shared space is nobody’s task. Each person’s reasonable assumption is that the other will handle it, because each has seen the other handle similar things. This is the whole problem in one sentence, and it gets worse the more competent both people are.

Divergent conventions. One names by date and vendor, the other by what it was for. Both are defensible. Together they produce an archive where searching requires knowing who filed the item, which is precisely the fact nobody records. See finding it again — a naming scheme only works if there’s one of it.

Asymmetric context. The reason for a spend is known by the person who made it and by nobody else. If they capture without annotating, the record is unreadable to their partner, who then can’t tell whether it’s already handled. The reason field is the one that doesn’t transfer, which is the argument for writing it down at the time applying twice as hard here.

Double capture. Both photograph the same receipt. Now there are two files, differently named, possibly in different places, and any later retrieval has to establish which is authoritative. Twice as much work for less certainty than one copy.

What actually resolves it

One destination. Not two, not per-person. Every capture goes to the same landing zone, whatever the medium, whoever did it. The moment there’s a “mine” and a “theirs”, the gap between them is invisible by construction. The two-container structure in an inbox and an archive is medium-independent and person-independent for the same reason.

An owner per class of record, not per item. Item-level ownership requires negotiating each time, which is friction that gets skipped. Class-level ownership is decided once: one person owns supplier invoices, the other owns anything paid on the joint card, one owns the vehicle, whatever fits. The rule you want is that every class has exactly one name against it, and no class has two.

A shared convention, written down. Not necessarily a good convention. A written mediocre one beats two unwritten good ones, because the value is in the agreement rather than the design. Half a page: how files are named, where they go, what the reason field looks like.

One person runs the periodic pass. The move from inbox to archive is a judgement-heavy job that produces inconsistency if it alternates. Same person each time; the other can do capture freely.

Say what you did, briefly. The whole of coordination overhead here is a short standing habit of mentioning the unusual ones — “I’ve put the plumber’s invoice in”, “the hotel one’s mine, it’s on the list”. Not a status meeting. Just enough that the diffusion assumption stops operating.

Keep or bin

KEEP OR BIN — a two-person system

  · One shared landing zone
                    → KEEP. Makes the gap visible
                      instead of structural.

  · A written naming convention, even a
    mediocre one
                    → KEEP. Agreement beats elegance.

  · One named owner per class of record
                    → KEEP. Item-level ownership needs
                      negotiating; class-level doesn't.

  · Two parallel personal archives
                    → BIN one, or merge them. The
                      overlap and the gap are both
                      unmeasurable.

  · Second photo of a receipt the other
    person already captured
                    → BIN. Two copies is a
                      verification problem, not
                      redundancy.

  · A capture with no reason attached,
    made by the other person
                    → NOT A USABLE RECORD to anyone
                      but them. Chase it while it's
                      recent.

  · Who is required to hold what, and
    whether one person's records satisfy
    an obligation on the other
                    → ASK LOCALLY. Depends on the
                      legal arrangement and the
                      jurisdiction.

Merging two archives that already exist

The common starting position, and it’s tractable in an afternoon if you accept an imperfect result.

Pick a target convention and don’t improve it. Whichever of the two is already applied to more material. Adopting the larger set costs less renaming.

Merge into one place before renaming anything. A single directory of badly-named files is a strictly better position than two directories of differently-named files, because you can at least see the whole set. Order matters: consolidate, then normalise.

Deduplicate by amount and date, not by filename. The two copies of a receipt will not share a name. They will share a total and a date, which is what makes a duplicate findable.

Resolve the reason gaps by asking, immediately. For anything recent, the other person still knows. This is the perishable part of the merge and the only part with a deadline.

Accept a labelled gap list. Items neither of you can identify go on a list saying so. As in reconstructing a year you didn’t record, a labelled gap is a defensible artefact and a confident guess is not.

When one person is doing the recordkeeping for both

A distinct arrangement and worth naming, because it’s extremely common and has its own failure: the non-filing person captures nothing and the filing person receives paper with no context attached.

The fix is that the capture obligation doesn’t move. Whoever spent it supplies the reason, in whatever form — a note on the slip, a message, a line on a list. The filer’s job is to file, and they cannot manufacture the one field the receipt never had.

There’s a records-hygiene point here too. Records about another person, even a partner or a colleague, carry more personal information than people expect — what a receipt reveals sets out what’s actually printed. Whether that carries any obligation depends on your situation and jurisdiction, and this site doesn’t say.

What this doesn’t settle

Who is required to keep which records in a partnership, a company, a household, or a contractor relationship. Whether one party holding a record discharges the other’s obligation. Whether records must be separated between personal and business, or between people.

Those are legal and tax questions, they turn on the arrangement between you and on where you are, and your adviser or tax authority answers them. What this page settles is the coordination problem underneath, which exists identically in every jurisdiction: two people with equal responsibility for a receipt produce, on average, slightly less than one person’s worth of filing.