HubSpot import: "Could not parse number"
HubSpot's own troubleshooting guide names “Could not parse number” as an import error, shown when a value in a number-mapped column carries non-numeric characters. HubSpot's own wording states the consequence directly: “The affected records were imported, but the number property does not contain a value” — the same silent pattern as a date that will not parse, on properties like Annual revenue and Number of employees that live on the company object, not the contact one.
Built for anyone preparing a contact or company file for HubSpot or Salesforce — sales operations, RevOps, and agencies migrating a client's data.
Read from HubSpot's own Knowledge BaseHubSpot Knowledge Base — knowledge.hubspot.com, read 13 September 2026.
What this error means
A HubSpot number property — Annual revenue and Number of employees on a company record are the two you will meet most — takes plain digits and nothing else. A cell holding “unknown,” a stray note, or a value HubSpot cannot read as a number at all is not rejected with the row: the company is created, and the property is simply left empty, the same silent loss HubSpot's own error text admits for a date that will not parse.
What causes it, and what our engine actually checks
CRM Import Check's numberNotNumeric rule reads every cell mapped to a number property and flags anything that is not a plain number and does not even look like a grouped one — the same cells HubSpot would otherwise import with the property left blank. A second rule, thousandsSeparator, catches a different and fixable case: a value like “1,250,000” that is a real number, just written with grouping HubSpot's own number field does not accept, and offers the plain digits it becomes once the separators are removed.
What the engine found in a sample file
In a three-row companies sample, Number of employees held “unknown” in one row and “N/A” in another — neither reads as a number at all — and the engine flagged both as numberNotNumeric, a blocker, the exact condition that leaves HubSpot's property blank on import. Annual revenue in the first row held “1,250,000”: a real figure, just grouped the way a spreadsheet writes it, which the engine flagged separately as thousandsSeparator, a warning, and offered “1250000” as the plain form HubSpot's own number field accepts.
numberNotNumericBlocker2 and 3“unknown”
thousandsSeparatorWarning1“1,250,000” → “1250000”
Blocker stops the import outright, or lets it through with data silently lost. Warning usually gets through but is worth checking before you trust the numbers. Notice is informational — safe to leave, safe to clear.
Reproduced from a sample file — read September 13, 2026.
Check your own file for this
Drop the company file you were about to import into HubSpot below. It is read in this browser tab, the same engine runs the same check, and you will see immediately whether it carries an unparseable number.
CSV, TSV or Excel, up to 20 MB
How to fix it by hand
For a cell that is not a number at all — “unknown,” “N/A,” a stray note — go back to the source the figure came from and fill it in, or leave the property blank on purpose rather than typing a placeholder HubSpot will store as if it meant something. For a cell that is a real number written with thousands separators, strip the commas or periods used for grouping before you import; the value underneath does not need correcting, only the punctuation around it.
Frequently asked questions
- What does HubSpot's "Could not parse number" error mean?
- It means a value in a number-mapped column — Annual revenue or Number of employees on a company, for instance — has characters HubSpot cannot read as a number at all. HubSpot does not reject the row: it imports the company anyway and leaves that property blank.
- Why does this page use a companies file instead of contacts?
- Because Annual revenue and Number of employees, the two properties this error most often shows up on, live on HubSpot's company object, not the contact one. Every other page in this family uses a contacts file; this is the first that has to use companies instead.
- Can CRM Import Check fix an unparseable number automatically?
- Only the fixable case: a real number written with thousands separators, where stripping the grouping punctuation is a mechanical, reversible fix. A value that is not a number at all — “unknown,” a stray note — is left for a person to fill in or knowingly leave blank.
What CRM Import Check does with this file
Upload a file carrying this problem into CRM Import Check and it lists every number cell it cannot read at all, by row, before you import anything — the exact cells HubSpot would otherwise leave blank. Where the value is only grouped rather than genuinely non-numeric, it offers the plain digits as a one-click fix, shown before and after, rather than lumping every unparseable cell into one undifferentiated warning.
The file you'd check is read in your browser tab — it is never uploaded anywhere.
What it cannot do here
The tool cannot supply a number that was never recorded anywhere — a company whose Annual revenue was never written down stays blank whatever the file says. It also does not know which of two different decimal or grouping conventions your organisation intends when a value could genuinely be read either way; it only fixes the grouping it can read unambiguously.