Somewhere in your directory there are probably two entries for the same family. Maybe the Garcias showed up as guests in September and got entered again as members in February, first name spelled two different ways. Maybe someone split a household when a daughter moved out for a semester and never merged it back. Whatever the cause, the result is the same: giving history split across two records, a care note sitting on the wrong one, and a directory that quietly lies to whoever pulls it up.
This is not a glamorous fix. There is no button that finds duplicates and folds them together while you get coffee. But the process is not hard either, and doing it carefully — in the right order, before the giving statements go out — is the difference between a directory people trust and one they quietly double-check.
How to actually spot a duplicate
The obvious cases are easy: same address, same phone number, a name spelled two ways. The harder ones look like this:
- A household entered once from a connect card at a guest table, and again months later when someone joined a small group and got added fresh instead of found.
- A woman entered under a maiden name before a wedding, and again under her married name after — two households, one person.
- A family that moved and got re-entered with the new address because the old one did not turn up in a quick search.
- Two entries for the same last name at the same address where one has three people on it and the other has one — often the same family, split by whoever entered the second card.
The pattern worth watching for is not identical data. It is plausible overlap: same phone number, same street, a first name in common, a child who shows up in both. When two records share two or more of those, stop and look closer before you assume they are unrelated.
Decide which record survives, before you touch anything
Pick the record with the most history attached — the one with giving entries, a care case, a serving assignment, group membership. That is your survivor. The other one is the record you are going to empty out and stop using. Write this decision down somewhere, even just a sticky note: “Garcia household, keep the one on Maple, retire the one on the connect card from September.” It sounds unnecessary until you are three steps into the process and cannot remember which was which.
If both records have real history — giving on one, a care case on the other — that is fine. You are not choosing the record with less to lose. You are choosing the one you will move everything onto, and then moving the rest to meet it there.
Move the people first
Every person profile belongs to a household. Before you touch giving or care, edit each person on the duplicate household and move them onto the surviving one. Once every person has been moved, the duplicate household should have nobody left on it — which is your signal that the hard part is done. If you are unclear on why the household is the unit that matters here rather than the individual, our piece on households vs. individuals walks through why SundayBridge tracks both and what each is for.
Re-attach the giving history
This is the step people are most afraid of, and the one most worth doing slowly. Every contribution recorded under the duplicate household needs to end up attached to the surviving one, under the right person, so a year-end statement reflects one total instead of two partial ones. Go entry by entry: edit each gift, confirm the giver and the amount, and re-save it against the correct person on the correct household. It is tedious for a household with a long giving history. It is also the one part of this process where a shortcut costs someone real money on a tax document, so resist the urge to eyeball totals instead of moving each entry.
If you are doing this cleanup in December, do it before you generate statements, not after. A statement printed from a split record undercounts a family’s giving for the year, and reprinting after the fact is a worse conversation than a few extra minutes now. For more on why giving records need this kind of care, see tracking giving that respects the giver.
Carry over care notes and serving history
Pastoral care cases and serving assignments are the two categories people forget, because they are not the first thing anyone thinks of when they hear the word “duplicate.” But a care case logged against the wrong household means the next person who opens that file sees nothing, and asks a family about a hospital stay nobody told them about, or misses one entirely. Move each open case and its comment history over to the person on the surviving household. Do the same for serving assignments and group membership — a volunteer who is on the usher rotation should not disappear from the schedule because their household got folded into another.
This is slower than the giving step, because there is no single field to double-check the way there is with a dollar amount. Read the case notes as you move them. It is worth five extra minutes to confirm nothing was lost in translation between one household and the other.
Retire the old record instead of trying to delete it
Household profiles can be created and edited, but there is no delete button for one that has already been folded into another. That is by design more than by accident: an empty household with no people, no giving, and no care attached to it simply stops appearing in anyone’s normal workflow. Leave it. Do not add anything new to it. If a volunteer stumbles across it later and asks why there is a household with nobody on it, the answer is simple: it used to be a duplicate, and everything on it has already moved.
If your directory has more than one or two of these sitting around from a spreadsheet import or a rushed setup, it is worth doing a full pass rather than fixing them one at a time as you notice them. Our guide to cleaning up a church database covers the broader version of this problem — duplicates, stale entries, people who moved away three years ago and never got marked inactive.
A concrete example of what is actually at stake
Say the Andersons give forty dollars a week and have for the whole year, but half those gifts landed on a household record created from a connect card back in March, and the other half landed on the older record from when they first joined a small group two years earlier. Fifty-two weeks at forty dollars is $2,080. If twenty-six of those weeks are attached to one record and twenty-six to the other, the statement you print for each record shows $1,040 — a real number, correctly calculated, and wrong for the family’s actual year. Nobody entered bad data. The data was just never in one place.
The same math applies in miniature to serving. If the Andersons are on the greeter rotation twice a month and half those assignments were logged against the duplicate household before it was created as a duplicate on purpose, whoever pulls up their serving history sees one assignment a month instead of two, and might reasonably conclude they are due for another rotation when they are not. Small errors like this do not look like errors. They look like a slightly wrong but plausible answer, which is worse, because nobody double-checks a plausible answer.
How to keep this from happening again
The root cause is almost always the same: whoever is entering a new family does not search first, or searches with a name that does not match what is already there. A quick habit fixes most of it — before creating a new household, search by phone number or street address, not just last name, since that is what actually catches the maiden-name and misspelling cases. If more than one person enters new households — a greeter on Sunday, a secretary during the week — make that search-first habit explicit rather than assumed. Our piece on building a church directory your team trusts goes into who should be entering data and how to keep one version of the truth instead of several.
None of this is glamorous work. But a directory where every gift, every care conversation, and every serving assignment sits on the right household is the difference between a system your staff reaches for without thinking, and one they quietly work around because they have stopped trusting what it shows them.