BillPad

My import fails, skips invoices or shows different totals

Resolve date, mapping, duplicate, total and capacity warnings in the import preview.

An import preview is a checkpoint before records change. Stop and compare it with the source file if dates, amounts or record counts look wrong. Keep the original export and an Export backup of your current BillPad workspace.

Try this first

  1. Use CSV or XLSX with column names in the first row.
  2. Check the file is within 5 MB and 5,000 rows.
  3. Inspect the selected data type and Match columns.
  4. Read skipped-row reasons and confirm any date-order question before importing.

Possible causes

03/04/2026 is ambiguous: it can mean 3 April or March 4. Missing invoice number, date, client or amount mapping can block useful records. XLSX reads only the first worksheet; Excel 1904-system dates need conversion to explicit dates in CSV.

Already imported invoices are skipped rather than overwritten. The 100-document beta limit includes existing documents; only the newest eligible invoices that fit are selected. An invoice above 100 lines is skipped whole. Archiving retains documents and does not create additional capacity.

Fix the problem

  1. Select DD/MM · 3 April or MM/DD · March 4 according to the source report.
  2. In Match columns, map invoice number, date and client, then total or the line rate/amount fields.
  3. Check currency, inclusive versus exclusive amounts, paid amounts and due dates.
  4. Compare several preview records with their originals before confirming.

Differences over one minor currency unit produce total warnings; the Total tile can flag even a one-unit difference. Mixed tax rates, negative lines and inferred payment dates also need review. Do not rename duplicates just to import them again: that can double your receivables.

For products and clients, Skip and Update apply to matched catalogue or contact records. Those policies do not mean invoice duplicates will be overwritten. Many vendor profiles remain provisional, so use manual mapping when the actual headings differ.

Still stuck

Email [email protected] with the source app, selected data type, exact warning and a redacted header/sample row. Keep the complete original file unchanged while the mapping is investigated.

Last updated 18 October 2026