Store orders may come from one export, advertising from another, and returns may arrive later. File import in ProfitVena Channel Economics helps organize observations and check their origin. A successful upload alone does not establish complete profit.
1. Keep identifiers stable
Order files need order and line IDs, dates, statuses, SKUs, quantities, and amounts. Cost files need entry IDs, dates, cost types, and amounts. Preserve IDs when re-exporting the same transaction so duplicate controls can work. Corrections use signed amounts and should be reconciled to the original entry. The import guide provides templates and field definitions.
2. Inspect the preview before committing
Upload CSV, TSV, or native .xlsx. Select the sheet and map columns where needed. Review rejected rows, possible duplicates, dates, return statuses, and amounts before explicit commit. A duplicate warning does not automatically erase a cost; compare the record to the independent source export.
3. Reconcile against a separate source report
After confirming the import, obtain a report from the store system for the same period. Compare row counts, quantities, and amounts at the relevant grouping level. Investigate missing, unmatched, and extra groups. A match on selected metrics at one point in time does not prove that taxes, returns, or all costs are complete. The data-lineage method explains what source checks do and do not establish.
4. Leave unknown values unknown
If the file contains orders and delivery costs but omits advertising or support, those expenses are unknown—not zero. A known-cost summary should not be described as fully loaded profit. Imported observations also do not silently replace a saved manual scenario; modeled assumptions and observed records remain distinct evidence bases. Use the channel-economics method when updating the comparison.
5. Separate funnel events from financial events
Cart and checkout activity describes the funnel, not recognized revenue. Connected sources and Store API events have separate workspace eligibility and source-acceptance requirements; an HTTP 202 means accepted for processing only. The event-recognition method sets out the boundary.
The practical sequence is simple: upload, inspect, commit, reconcile independently, and label the output accurately—as a scenario, known-cost summary, or reconciled observation. Then the channel decision rests on evidence you actually have.
