Distributors outgrowing QuickBooks are the most common new Prophet 21 buyer, and the migration path is the least documented thing about it. There is no connector and no third-party tool that does this well. You extract, you map, and you load.
That is fine. It is a known shape of work. What sinks projects is treating it as a data-copy exercise rather than a design exercise with a data step at the end.
What actually comes across
The useful framing is that QuickBooks holds four things you want and one you mostly do not.
Customers. Straightforward. The trap is ship-to addresses: QuickBooks often carries them as free text on the transaction, while P21 wants them as structured ship-to records under the customer. That is a modelling decision someone has to make, not a column mapping.
Vendors and items. Item data out of QuickBooks is usually thinner than P21 needs. There is no supplier-by-location link, units of measure are informal, and product grouping frequently does not exist in any structured way. Expect to author data that has no source, not just move data that does.
Open AR and AP. These come from the aging detail reports, not from the transaction tables. Load them as open documents at cutover, not as history.
GL opening balances. From the trial balance, as a journal entry, on the cutover date. This is the piece that must tie exactly.
History. Mostly leave it. Re-booking years of invoices through order entry to recreate a sales history is how migrations lose their timeline. Keep the QuickBooks file readable for lookups and bring across only what the business genuinely needs to transact against.
The order
- Chart of accounts designed in P21 terms, then company GL defaults set
- Code tables: terms, tax groups, product groups, units of measure
- Vendors, then customers
- Items, with their unit and supplier links complete
- Freeze QuickBooks entry
- Open AR, open AP, open orders, open POs
- GL opening balances from the trial balance
- Reconcile
- Go live
Steps 5 through 9 are a cutover weekend, not a phase. Everything before them is rehearsable and should be rehearsed.
The condensing problem
QuickBooks files that have never been condensed get very large, and the export step is slower and more fragile than anyone plans for. Find out early how big the file is and how far back it goes, because it changes the extract approach and it is the sort of thing that is discovered at the worst moment.
Where the arguments happen
Two decisions cause most of the friction, and both are worth forcing early:
Item numbering. Clients frequently want to adopt a new convention during the move. That is legitimate, and it means the legacy and new catalogs share no identifier. If that is the choice, the cross-reference between old and new is a deliverable in its own right, built and reconciled before pricing work starts.
How much history. Everyone asks for all of it. Almost nobody needs it. Settle this in writing, because it is the single biggest lever on the size of the project, and an unspoken assumption here is what turns a quoted migration into an over-run one.
When it is done
Not when the loads finish. A QuickBooks migration is done when:
- the trial balance in P21 matches the trial balance in QuickBooks on the cutover date
- open AR totals match the aging detail
- open AP totals match the aging detail
- item and customer counts reconcile against the source with the exclusions documented
Put those four checks in the plan as an explicit stage with a name and a sign-off. If the reconciliation is not a scheduled step, it becomes something that happens under pressure on the Monday after go-live, or it does not happen at all.