Auto-Match and Matching Settings
Reviewing every proposed match is fine for a few hundred Accounts and tedious for a few thousand. Auto-match confirms the unambiguous ones; Settings controls what counts as a match in the first place.
Auto-match
After analysis, the Auto-match button in the analysis bar shows how many rows qualify. Click it and choose:
- High confidence — rows scoring 80 or more: a name match backed by at least one other field.
- High + Medium — also rows scoring 55 to 79: a name match on its own.
- Also create a new Account for no-match Xero contacts — optionally, rows with no suggestion get a new Account created and linked.
Auto-match only ever applies rows with a single suggestion. Needs Review rows are left alone, however high their scores. Every link it makes is flagged so the whole run can be reverted with Undo last auto-matched run under Automation.
Organisations with related entities — "Acme Pty Ltd", "Acme Holdings", "Acme (NZ)" — are exactly where automatic matching goes wrong. The tool routes most of these to Needs Review, but if your customer list has many lookalikes, run High confidence only and review Medium by hand.
Settings
Settings in the toolbar opens the matching method for the current profile.
Smart matching
The default and recommended method. It scores candidates by how similar they are, tolerating typos, abbreviations and word order.
- Main field to match on finds candidates. For Accounts it's Xero Customer Name against Account Name, compared as similar text. Ticking This field identifies which record, but doesn't score on its own turns it into a gate: it narrows the candidates but the extra fields decide the confidence. That's how the Invoice profile works, where a customer name finds the Opportunities and the amount decides.
- Extra fields that raise confidence are optional pairs that add to the score when they agree. The defaults for Accounts are account number, street, city, postcode, state and phone. Add, remove or change them with the dropdowns.
- Required on an extra field means a match can't be trusted enough to auto-apply unless that field agrees. Invoices require the amount.
The pre-set signals suit most orgs. Change them only to fine-tune: for example, adding an ABN field you've captured on both sides as an exact, required signal.
Rule-based matching
Replaces scoring with exact field rules you define, combined with all or any logic, or an advanced expression. A record matches only when the rules are met, so there are no Medium or Low results, only match or no match. It suits orgs with a reliable shared key. Purchase Orders use rules only, since a purchase order has no name to compare against an Opportunity.
Switching method changes how suggestions are found from the next analysis on. It doesn't touch existing links.
A wrong link is cheap to fix before any documents exist and expensive afterwards, once invoices have been raised against the wrong Xero contact. After a big auto-match run, spot-check a sample from the Matched filter.
Xero
Real Estate
Equifax