Migrating Off QuickBooks: The Full Checklist

The data transfer takes a few hours. The migration takes two to three weeks.

That gap is the entire subject, and misjudging it is why migrations that should be routine turn into months of nobody trusting the numbers.

Everything below is ordered so the expensive discoveries happen in week one rather than week five.

The short answer

Week one decide and check, week two clean, week three migrate, weeks four and five validate and reconnect. Do not compress it, do not do it in your busy season, and do not cancel the old subscription until the new system has survived a full cycle.

Step 0: the check that changes the plan

Before anything else, find out how large your company file is.

Files exceeding roughly 750,000 targets cannot be converted to QuickBooks Online. Above that threshold, migration is not a conversion. You export lists and balances, start fresh, and keep the old file as a historical archive.

If you are over the limit, timing stops being flexible. Starting fresh mid-year leaves you with a partial year in each system and comparatives that live nowhere convenient, so the sensible plan may be to wait for the next year start. That is a decision you can only make if you check now.

Week 1: decide and prepare

Confirm what you are migrating to. If this is not settled, settle it before touching data. For contractors in particular the destination is often not another accounting product at all, which is covered in QuickBooks alternatives for contractors.

Back up the company file. Independently, to storage you control, not relying on any vendor's process.

Export a full archive while you still have complete access. Reports as PDF for every year, lists and transactions as CSV, the company file itself if on Desktop. This archive is worth having regardless of how the migration goes, and it is the thing you cannot recreate later.

List every integration touching QuickBooks. Payroll, payments, time tracking, CRM, bank feeds, e-commerce, expense apps. Write them down with who administers each. Every one is a separate reconnection task in week five.

Book your accountant now. Their capacity disappears in the months approaching each support deadline, and you want them for both the cleanup and the validation.

Pick the date. Period end, ideally financial year start, comfortably outside your busy season. For seasonal trades this dominates everything else, which is the same logic as the advertising calendar in the best time of year to advertise HVAC.

Week 2: clean up

The step everyone wants to skip and the one that determines whether this goes well.

• Merge duplicate customers and vendors. Migrating duplicates carries them forward and multiplies them at the first CRM sync.
• Standardise naming. Ltd versus Limited, ampersands, trading names. Pick a convention and apply it.
• Close inactive accounts and items you have not used in two years.
• Resolve outstanding reconciliation items. Anything unreconciled becomes a mystery in the new system that nobody can trace.
• Clear undeposited funds. A classic source of post-migration discrepancy.
• Review the chart of accounts. A migration is the one good moment to simplify it, and there will not be another for years.
• Write off genuinely uncollectable receivables rather than carrying them across.

Week 3: migrate

Run the transfer. Follow the destination vendor's process; they all differ.

Do not touch the old file afterwards. It is your reference copy and your only way to answer "what did this look like before."

Record the cutover date and time precisely. Every validation check below is relative to it.

Do not let anyone enter new transactions in either system until validation is done. A day of frozen entry is inconvenient. Reconciling transactions entered into both systems during an unvalidated window is genuinely painful.

Week 4: validate

The checks, in order. Each compares the new system against the old file at the cutover date. Every difference must be explained before you proceed, not noted and moved past.

1. Trial balance. Must match exactly.
2. Profit and loss for each prior year.
3. Balance sheet at the cutover date.
4. Accounts receivable ageing, total and by customer.
5. Accounts payable ageing, same.
6. Bank and credit card balances at cutover.
7. Open invoice count and total.
8. Customer and vendor counts.
9. Inventory quantities and values, if you carry stock.
10. Payroll year-to-date figures per employee, if payroll moved.

Then reproduce the five reports you actually run every month. Not a sample; the specific five. This is where migration regret concentrates: custom reports built up over years often have no direct equivalent, and finding that out in week four is fixable while finding out in month three is not.

Have someone other than the person who ran the migration sign off on this. Self-checking a migration is how differences get rationalised.

Week 5: reconnect and go live

Reconnect integrations one at a time. Verify each before starting the next. Bundling them means a failure you cannot attribute.

Bank feeds first, since everything else depends on transactions flowing.

Payroll next, tested on a single employee before running a full cycle.

The CRM last, and deliberately. A first sync into a freshly migrated file is the single most likely moment to create duplicate customers at scale. Confirm the matching field, set it to one-way at quote acceptance, sync ten records and inspect both systems by hand before doing more. The failure modes are in QuickBooks CRM sync problems.

Run parallel for one cycle. One month with a period end in it. Not longer: dual entry diverges, nobody maintains both, and you end up trusting neither.

Then cancel, with your exports already in hand. Read-only access to QuickBooks lasts one year and then ends, and Payroll and Payments often cancel separately. The full process is in how to cancel a QuickBooks subscription.

Who does what

Migrations stall when everyone assumes someone else owns a step. Assign these explicitly, in writing, before week one.

The decision owner. Usually the business owner. Picks the destination, sets the date, and is the person who says go or no-go after validation. One person, not a committee.

The data owner. Whoever knows the file best, normally the bookkeeper. Runs the cleanup in week two and is the single most important person in the project. If this is an external bookkeeper, book their time early and budget for more hours than the transfer itself suggests.

The accountant. Advises on destination, assists with cleanup, and independently signs off validation. Deliberately not the same person who ran the migration.

The integration owner. Whoever administers each connected system. Frequently several people, which is exactly why the week-one inventory matters.

The users. Everyone entering data daily. They need the new workflow before go-live, and they need to have touched it during the trial. This is the group most often left until the week of the switch, and their adoption decides whether the new system holds.

One meeting in week one with all of them, one in week four to review validation, and a short daily check-in through week five is enough coordination for a project this size. More than that becomes theatre.

A go or no-go test before you commit

Before cancelling anything or letting the old system go cold, five questions. All must be yes.

1. Does the trial balance match the old file at the cutover date, exactly?
2. Can you reproduce the five reports you actually run every month?
3. Have all integrations been reconnected and verified individually, including a real payroll run and a real bank feed?
4. Has someone other than the migrator signed off the validation?
5. Has the team entered a full week of real work into the new system without falling back to the old one?

Any no means you are not finished, and the correct response is to keep both systems running for another cycle rather than pushing forward. That is a modest cost. Discovering a gap after the old subscription has lapsed and the read-only clock has started is a considerably larger one.

Question five catches something the others cannot. A migration can be technically flawless and still fail, because people quietly keep working the old way and the new system slowly fills with gaps. If the team is still reaching for the old file in week five, the problem is workflow fit rather than data, and it is better to find that out while you still have both.

The mistakes that cost the most

Migrating against a deadline. Every accountant and consultant is handling the same rush in the same months. Availability falls, rates rise, and you have no slack if something needs a second attempt.

Migrating in busy season. A fortnight of uncertain numbers is survivable in a quiet month and damaging in a peak one.

Skipping cleanup. The largest source of post-migration pain and entirely within your control.

Assuming integrations reconnect themselves. They do not, and several fail silently.

Cancelling too early. Removes your comparison source at the moment you most need it.

Treating it as purely technical. The people entering data need the new workflow before go-live, not during it. Most failed migrations are adoption failures with a technical postmortem attached.

Recreating every old habit. Some of your current routines exist to work around limitations that no longer apply. Spend an hour identifying which, because dropping them is a bigger gain than anything in the feature comparison.

If you are on Desktop, the clock is real

Support for QuickBooks Desktop 2023 ends 31 May 2026. Desktop 2024 is the final version, supported through 30 September 2027. There will be no further releases, and new Pro Plus, Premier Plus and Mac Plus subscriptions stopped selling on 30 September 2024.

Working backwards from those dates with a three-week project, an accountant to book, and a preference for a quiet period end, the sensible start date is considerably earlier than the deadline suggests. The detail is in QuickBooks Desktop discontinued, and the Online comparison with the end date factored in is in QuickBooks Online versus Desktop in 2026.

One thing to fix while you are in here

A migration moves your accounting data. It does not create the records that were never captured in the first place.

QuickBooks holds people who paid you. It holds nothing about enquiries that never converted, quotes that went quiet, or where a customer came from, which means it cannot produce a close rate, lead source performance or cost per booked job. Neither can the product you are moving to, because they all begin at the invoice.

If you are rebuilding your systems anyway, this is the moment to close that gap rather than carrying it forward for another decade. The reasoning is in is QuickBooks a CRM, the integration boundary that keeps the two cleanly separated is in the best CRM that integrates with QuickBooks, and the scale of what those missing records cost is in how many HVAC leads you need per month: at a moderate job volume, a five-point close rate improvement is worth over $130,000 a year, and it is invisible without them.