Getting started

Why you should clean your data before you migrate it, not after

The dead phone number is a five-second fix today. In a new system, it is a mystery.

8 min read

Ask AI · in the $19/mo plan

Ask your own records a question. Which regulars have quietly stopped coming?” — answered from the records you already keep. It reads your church and no other, and it can't invent a number.

10 questions a month included · no AI add-on to buy

Every church spreadsheet has a Bob Smith problem. There are two of him, entered eight months apart by two different volunteers, and nobody remembers which one has the current cell number. Multiply that by a decade of well-meaning data entry and you get a file that looks complete but is quietly unreliable in a dozen small ways.

The instinct, when you finally decide to leave that spreadsheet behind, is to move fast: copy everything into the new system, then tidy it up once it is somewhere better. This is backwards, and it is the single most common reason a data migration disappoints people. Clean first. Move second. Here is why the order matters more than it seems like it should.

A mess in a spreadsheet is a five-minute fix

Right now, today, fixing a duplicate record takes almost no effort. You can see both rows side by side, in the same font, on the same screen. You can tell at a glance that “Bob Smith” and “Robert Smith, 412 Oak St.” are the same household because you already know the family. Deleting one row and keeping the better one is a five-second decision.

That same fix, made after the move, is not five seconds. It is a search, a comparison across two screens or two exports, a moment of doubt about which record has the correct giving history attached, and often a decision deferred because nobody wants to be the one who deletes the wrong Bob Smith. The work does not get easier by waiting. It gets harder, because the context that made the decision obvious — the shared knowledge in your head, the surrounding cells, the years of muscle memory about who is who — does not travel with the data. It stays behind in the old file, and then the old file gets closed for the last time.

Duplicates cost you twice, not once

A duplicate record is not neutral. It costs you once when someone has to figure out which entry is correct, and it costs you again every time it is used before that happens. Two records for the same giver split their giving history in half, which understates their total on paper. Two records for the same volunteer mean a background-check note attached to one and a serving assignment attached to the other, so neither looks complete. A follow-up note left on the wrong Bob Smith is a note nobody will ever read.

None of that is a software problem. It is a data problem, and software inherits whatever you hand it. A new system with a duplicate in it is not cleaner than the old spreadsheet with the same duplicate — it just presents the same error with a nicer font. If you are still weighing whether to make the move at all, our guide to choosing church management software is worth reading first, because the honest answer is that no tool fixes a dirty list for you.

Dead contact information does not announce itself

A wrong phone number does not raise its hand. It sits in the directory looking exactly like a correct one, and the only way anyone finds out is when someone dials it and gets a stranger, or a disconnected line, or silence. In the old spreadsheet, you have a chance to catch some of these before the move: an address that bounced back mail last year, a number a volunteer mentioned was wrong, a household you know moved out of state. Those are cheap to fix now because you or someone on your team still remembers the detail that flags them.

Once the record is sitting in a new system, treated as current because it exists there at all, that flag disappears. Nobody double-checks an entry just because it recently moved house. The bad number becomes indistinguishable from a good one until a volunteer wastes a Saturday afternoon calling it.

Decide who is active before you decide anything else

Most church spreadsheets accumulate people the way a garage accumulates boxes: everyone who was ever added is still there, whether or not they are still around. Before you migrate anything, go through the list and mark who is genuinely active, who has drifted, and who left years ago and should not appear on a current roster at all. This single pass does more for the trustworthiness of your eventual directory than almost anything else you can do.

This is also the moment to settle how you group people, since it changes how the rest of the cleanup goes. If you have not already decided whether your directory should be organized by family or by person, read households versus individuals before you start marking anyone active or inactive — it is much easier to apply that decision once, to a spreadsheet, than to restructure a hundred records after they have already moved.

Marking someone inactive is not the same as deleting them. Keep the record and its history; just stop treating it as current. A background-check note, a giving history, or a serving assignment from three years ago is still real information about your church, even for someone who is no longer part of it.

Standardize the small things while you can still see them all

Phone numbers written four different ways, email addresses with stray capital letters, birthdays entered as text in one row and a date in another — none of this matters much while everything lives in one flexible spreadsheet, because a human reads every cell and makes sense of it. It matters more the moment those fields need to sort, filter, or export consistently, which is exactly what a real system will ask of them.

  • Pick one phone format and apply it everywhere — even a simple find-and-replace pass catches most of it.
  • Lowercase and trim stray spaces from email addresses; a trailing space turns a working address into a bounce.
  • Settle on one date format for birthdays and anniversaries before you type a single one into a new place.
  • Note who is a child versus an adult now, while you still remember, rather than guessing later from a birth year.

None of this is glamorous work, and none of it will show up in a demo. It shows up six months later, in whether the reports and exports you rely on actually mean what they say.

Giving history deserves the most care of anyone

Of everything in the spreadsheet, giving is the one place where a small error becomes a real problem, not just an annoyance. A person who gave through two duplicate records will see an incomplete total on their year-end statement. A typo in a contribution amount is a number someone may need to defend later. Before you move giving anywhere, reconcile the spreadsheet against your actual bank deposits so the totals you carry forward are the totals that actually happened, not the totals someone meant to type.

Do this reconciliation in the old file, where the deposit records and the spreadsheet rows are both still easy to lay side by side. Once giving is entered into a new system, correcting it means finding the specific contribution, editing it, and trusting that nothing downstream — a statement, a total, a pledge — already used the wrong figure. SundayBridge lets you correct a contribution after the fact, but a wrong number that already went into a printed statement is a harder conversation than a wrong number caught in a spreadsheet before anyone saw it.

A clean file makes the actual move almost boring

The payoff for all of this is that the migration itself becomes the easy part. Typing or importing clean, deduplicated, current records into a new home is mechanical work — there is no judgment left to exercise, because you already made every hard call in the spreadsheet. Our guide on moving your church off spreadsheets covers that mechanical part in full, including what to bring over first and how to bring your volunteers along. It assumes, though, that the file you are moving from is one you would actually vouch for. If it is not there yet, our database cleanup guide walks through the dedupe-and-verify process step by step, at whatever pace an evening or two allows.

A new system does not know your congregation. It only knows what you tell it, which means the trust people place in it on day one is entirely borrowed from how much you trusted the file you handed over. Spend the time before the move, not after. It is the same work either way — it is just far cheaper while the mess is still one you understand.

If you only fix one thing before you migrate, fix duplicates. A duplicate person quietly splits everything attached to them — giving, attendance, follow-up — in half, and nothing about a new system will notice or repair that for you.

Frequently asked questions

How long does cleaning up a church spreadsheet actually take?
For a church of 60 to 250 people, plan on one focused evening for the directory and another for giving history. Sorting by last name to find duplicates, checking a handful of phone numbers, and marking who is inactive is slow but not hard work. Spread over two sittings, most small churches finish it in under six hours total.
What if we do not have time to clean everything before we move?
Clean the parts that will embarrass you first: duplicate names, giving totals, and anyone marked active who moved away years ago. A partly clean directory that keeps improving is fine. A fully messy one that everyone has quietly agreed to stop trusting is the outcome to avoid, and it happens fast once people stop believing the data.
Should we delete the people who left instead of just marking them inactive?
No. Deleting erases attendance and giving history that may matter later — for a year-end statement, a background-check record, or simply remembering who used to be part of your church. Mark records inactive rather than removing them. A church database is a memory, and memories should not be edited away just because someone stopped attending.
Who should actually do the cleaning?
One person, not a committee. Cleaning a spreadsheet involves a hundred small judgment calls — is this the same Bob Smith, is this address current — and those calls need to be made consistently by someone who knows the congregation. A second person can spot-check the finished file, but the actual editing goes faster and stays more consistent with one owner.