Services
Monthly bookkeeping run entirely in Xero — closed, reviewed and proven every period by the team that does the work.
Bookkeeping packages
The Xero Handbook
Setting up Xero takes an afternoon. Setting it up so it still makes sense in three years takes longer, and the decisions that matter are made early. This is what those decisions are, in the order they arise, and which ones are expensive to reverse.
First
A migration copies your accounting into a new system. It does not improve it. Whatever is wrong in the old file arrives intact in the new one, usually harder to see because nobody recognises the layout yet.
So the first question is not how to move, it is what state the source is in. Three things are worth establishing before any data moves:
If the answers are unclear, clean up first and migrate second. It feels slower and it is faster. Migrating a file you do not trust means doing the cleanup twice, once in a system you know and once in a system you do not.
Decision one
The conversion date is the point where the old system stops being the record and Xero starts. Everything before it comes across as balances; everything after it is entered in Xero.
The start of a financial year is the cleanest choice. Comparatives are whole years, your tax preparer works from one system for the year, and there is no split period to explain. If you are within a couple of months of year end, waiting is usually worth it.
The start of a quarter is the practical compromise. Common, workable, and the reason most migrations happen when they do. It means one year of accounts spans two systems, which is manageable as long as everyone knows.
Mid-month is a decision to avoid. Partial periods make reconciliation ambiguous and every comparison needs a footnote.
Decision two
This is the decision with the longest shadow. The chart of accounts determines what questions your reports can answer, and changing it later means recoding history or losing comparability.
The most common mistake is importing the old chart unchanged because it is easier. The second most common is the opposite: rebuilding it from scratch so nothing reconciles to prior years. What works is a deliberate middle course.

Decision three
Conversion balances are the opening position on the day Xero takes over: what you own, what you owe, and the accumulated result of everything before. They are entered once and they anchor every report that follows.
Two rules make this straightforward. They must come from a set of accounts someone has signed off, ideally the last filed or reviewed period rather than a live figure that is still moving. And they must balance: assets equal liabilities plus equity, with the difference carried in retained earnings rather than an equity plug that nobody can explain later.
Outstanding invoices and bills are entered individually, not as a single receivables or payables total, because the aging has to work from day one. That means listing every unpaid invoice and bill at the conversion date, which is tedious and not optional.
If a historical adjustment appears after go-live, resist the urge to post it against conversion balances. Post it in the period it belongs to, or as a documented current-period adjustment. Editing opening balances after the fact is how a file stops agreeing with the accounts it was built from.
Decision four
Xero does not require historical detail to work. The question is what you lose without it, and the answer depends on who needs to look back.
Whatever you choose, keep the old system readable for the statutory retention period. Migrating is not archiving, and a cancelled subscription can take your history with it.
Decision five
Xero rarely sits alone. Bank feeds, a payroll provider, receipt capture, and whatever runs sales or inventory all have to connect, and the order matters.
Bank feeds first, because they take longest to establish and some connections need bank-side authorisation. Set them up before go-live so the first reconciliation is not also a troubleshooting exercise.
One system at a time after that. Connect, post a few transactions, check they land where you expect, then move on. Connecting five apps at once and discovering the numbers are wrong tells you nothing about which one is responsible.
Decide what posts in detail and what posts as a summary. A high-volume sales channel posting every line will make the file slow and the reconciliation unreadable. Daily totals with the detail held in the source system is usually the better design, and it is much harder to change later.
The expensive ones
Most setup choices can be changed in an afternoon. These cannot, and they are worth deciding slowly.
Go-live
The last step is the one people skip. Before the old system is switched off, run one period where the numbers are checked against something known: reconcile every bank and card account to a statement, agree the receivables and payables totals to the aging reports, and check the trial balance against the closing position from the old system.
If those three agree, the migration worked. If they do not, the gap is findable now and much harder to find in six months.
Then set your lock dates. A migrated file with no lock date invites someone to post into the conversion period, and that is how an opening balance quietly stops matching the accounts it came from.
Questions
Next
Tell us what you are on now and what state it is in. We will tell you whether to clean up first, what the migration involves, and what each part would cost.
Talk about a migration See the 32-point check