Automating Excel RFQs without losing line-item context

A customer’s Excel RFQ may be ready to import, or it may contain several tabs, notes above the table and part descriptions such as “same as last shutdown.” Those are different kinds of work.

For consistent columns and exact item numbers, a validated import can remove most of the copying. Bourne is a better fit when your reps need to interpret the workbook against customer history and product information before they can quote. This guide explains how to tell the difference and preserve every requested line through the process.

Use a conventional import when the structure is stable

Microsoft’s Power Query combine-files operation applies transformations to files with the same structure. Its folder connector uses a sample-file transformation for the other files, so check table names and columns before treating the result as complete.

A customer’s free-form notes or a changed layout can still require interpretation. Keep import failures visible rather than silently dropping rows that do not fit the template.

Keep the original row beside the proposed quote line

Keep the original workbook, sheet name and row reference. Separate raw values from resolved values. A rep needs to know whether “12” came from a quantity cell, a pack description or an earlier order.

Use a review table such as this example:

Keep the buyer’s description beside the internal SKU. When you resolve “same as last shutdown” through quote history, retain the earlier quote line and the equipment reference that support the match. The reviewer can then check whether the history applies to this request.

Check the selling unit before applying the price

A correct part number can still produce a wrong quote if the selling unit differs from the requested unit. For example, a request for 600 metres is not the same as 600 reels. Confirm both the quantity and the basis of the price.

Keep customer discounts, validity dates and currency separate from historical unit prices. Odoo’s pricelist documentation illustrates why dates and quantity conditions matter in current pricing. The same review principle applies even if your prices live elsewhere.

Compare revised workbooks by content, even when the filename is unchanged

Suppose the buyer sends RFQ-final.xlsx and later sends another file with the same name. Compare content and the buyer’s instructions, not filenames alone. Mark the current revision and retain the earlier one.

Check whether the new sheet changes existing rows or adds more work. Keep the previous ERP quote identifier with the request so the automation can prepare an amendment instead of a duplicate. For files arriving through Gmail, retain the thread and message identifiers with the revision. In a shared Outlook inbox, also keep a named owner on the request so another rep does not process the revised attachment as new work.

A workbook revision can also change the engineering input behind a row. ASME Y14.35 covers how teams identify and record engineering drawing revisions. Keep the drawing number and revision as separate fields alongside the commercial RFQ version, rather than treating the latest spreadsheet as the complete specification.

For example, the buyer may leave the item description and quantity unchanged but replace drawing C with D in the attachments. A comparison based only on descriptions and quantities would report no change. Flag the revised drawing against the affected lines, retain the earlier estimate and ask the estimator which operations or inspection requirements need a new cost.

Test the workbook cases your team actually receives

Include multiple sheets, merged cells, formulas, hidden rows and notes outside the main table. Decide explicitly whether hidden rows represent excluded work or information the buyer expects you to use. That decision depends on the file and the commercial request.

The pilot acceptance criteria should cover line completeness as well as matching accuracy. Did every requested line appear in the draft, including lines the agent could not resolve? An omitted line and an incorrect match need separate results.

Choose Bourne when the workbook needs interpretation

Use Power Query for stable columns and repeatable transformations. Choose Bourne when your customers send different layouts, revise quantities across tabs or refer to parts by their own names. We can build the workflow that interprets those workbooks against your catalog and customer records, with the original cells beside the proposed quote lines.

Your team gets a quote to review instead of a cleaner spreadsheet to retype. Build Excel RFQ automation with Bourne.

Frequently asked questions

Do we need AI if customers already use our Excel template?

Possibly not. Start with a validated import when the template and item identifiers are reliable. Add interpretation only for work that the import cannot complete, such as customer aliases or instructions in notes.

Should we convert an Excel file to PDF before processing it?

Keep the original workbook available. Converting it can remove useful structure such as sheet names, formulas and cell types. A PDF can be a review aid, but should not be the only source when the spreadsheet contains relevant data.

Can the agent quote from a customer’s part numbers?

Yes, if you provide account-specific mappings or reliable history. Keep the original number next to the proposed internal item and route ambiguous matches to review.

How do we avoid dropping unmatched rows?

Require the output to account for every source line. An unresolved line should carry a status and a question, not disappear from the quotation queue. Reconcile input and output line counts during testing.

Arda Bulut

Arda Bulut is the co-founder and CTO of Bourne and HockeyStack. He leads engineering at Bourne, building the platform people use to create AI products, agents and automations.