E-commerce
Analyze return rates and reasons from your shop data
Upload one CSV or XLSX that already has order and return counts per product. You get return rates and grouped return reasons in a workbook.
or drop it here
- CSV / XLSX
- Up to 50 rows
- 64,000 extracted characters maximum
SKU-A | orders=100 | returns=8 | reason=too small SKU-B | orders=50 | returns=2 | reason=damaged box
- SKU-A counts: 100 orders / 8 returns
- Reason example: too small
Return-rate workbook, reason-group CSV, evidence findings and review worksheet
One clear job, from source to download
- 1
Add the source
Supported formats and limits are visible before the upload.
- 2
Confirm the settings
Review the exact source, options, units and access before processing.
- 3
Inspect and download
Check the preview and warnings, then unlock the complete package.
Return rates and grouped reasons from one table
One table with orders and returns per product
You upload one CSV or XLSX file with up to 50 rows and 64,000 extracted characters. Each row must already carry the order count and return count for a product, plus the return reason text. The tool does not join a separate orders export with a returns export, which keeps every rate traceable to a row. Filter or split larger exports first. This AI tool needs an account and costs 2 credits per job, shown before you start.
The four files you get back
You download a return-rate workbook, a reason-group CSV, a findings file with evidence and a review worksheet. Rates are calculated from the counts in your rows, such as 8 returns on 100 orders for one SKU. Reasons are grouped with real examples from your data, such as too small or damaged box. Check a few rates against your own shop figures before you share the workbook.
Sample size, language and cause
Minimum orders per finding sets the sample size below which a product is not reported. Three orders with one return give a 33 percent rate that says very little, so thin rows are held back. Reason language can be Auto, English or German, to match how your customers wrote. The groups show patterns in your data, not proven causes. Finding out why an item comes back stays your job.
Questions before you run it
Does the tool join my orders export with my returns export?
No. Your table has to already carry order counts and return counts on the same row. Nothing is joined behind your back, which is what keeps every rate traceable to a row.
How does the minimum orders per finding setting work?
It sets the sample size below which a SKU is not reported as a finding. Three orders and one return produce a 33 percent rate that means very little, so thin rows are held back instead of published as a signal.
Will it tell me why customers are returning an item?
It groups the reason text you supplied and shows source examples, such as too small or damaged box. Those groupings are patterns in your data, not established causes.
Which formats and how many rows can I upload?
CSV and XLSX, up to 50 rows and 64,000 extracted characters. Larger exports need filtering or splitting first.
Can I run a German-language returns export?
Yes. Reason language can be set to Auto, English or German, so the grouping follows the wording your customers actually used.