Data-health monitoring for practices: flags messy Xero/QuickBooks data across a client book. Here's exactly where Dext Precision and Postlock overlap - and where they don't.
A live accuracy score per client, catching unreconciled transactions, duplicate contacts, missing bills and wrong VAT codes - genuinely useful hygiene monitoring across a whole practice book.
Precision asks "is this client's data clean?" - a data-quality question. Postlock asks "did a number I already signed off and reported change?" - a sign-off-integrity question. A practice could reasonably run both; they don't overlap.
| Feature | Postlock | Dext Precision |
|---|---|---|
| Post-lock change detection (auto watchdog) | - | |
| Real Excel export · working-paper quality | - | |
| Per-line institutional memory (carry-forward notes) | - | |
| Supporting schedules | - | |
| Board-figure tracking | - | |
| Read-only Xero · no write access | ||
| Who-changed-it attribution | - | |
| Audit trail + adjustment register | - | |
| Named-user roles + approval workflow | ||
| Practice multi-client view | ||
| Price (early access) |
"?" = not confirmed from public docs; check Dext Precision's current site before deciding.
Post-lock monitoring: we haven't found another tool that watches closed periods and names the document and person behind a change - check us on it in the table below. The Excel pack: the tools we've compared keep reconciliations in their own dashboards; Postlock exports a real file because auditors and accountants live in files, not dashboards.
If you need multi-entity consolidation, driver-based forecasting, deep workflow, or client data-hygiene monitoring, look at ScaleXP, Numeric or Dext - they go further in those specific areas. Postlock is deliberate and narrow.
We request only read-only Xero access - technically incapable of changing your ledger. If Postlock were ever compromised, an attacker could see but never touch your numbers. That's a trust decision we won't trade away.
Every check here is deterministic and re-performable - the same inputs give the same answer, which is what audit evidence requires. Your ledger is never read by a model, and there's no metered AI cost under the price. If a tool's pitch leans on AI reading your transactions, ask two questions: what does it find that a written rule couldn't, and what happens to their costs when model pricing moves. Our full reasoning →
See how Postlock compares to the rest of the market on the full comparison hub.