Guide · Inside sales
RFQ automation for technical distributors: from email to an approved quote
A request for quotation with forty lines is not hard because of the forty lines. It is hard because every line is written in the customer's words: their own part numbers, a competitor's type code, “3 boxes” where you sell by the piece. This guide walks through each step between an incoming RFQ and a quote you can send — what can be automated safely, what must stay with a person, and what to test before you buy any tool for it.
Updated September 28, 2026
Where the time goes in an RFQ
Typing the lines into the ERP is the visible part. The expensive part is the search in between: finding which of your articles the customer means, checking whether the quantity is pieces or packs, looking up the customer's conditions and noticing that an item has been discontinued. An experienced inside sales rep does this from memory; a new one does it by asking colleagues.
Automation helps exactly where this search follows rules — and becomes dangerous where it guesses. The useful question for any tool is therefore not “how much does it match?” but “what does it do when it is not sure?”.
The steps from inbox to quote
- Intake — email, PDF, Excel or CSV. The customer is recognised by the sender's domain, not only the address.
- Extraction — header data (deadline, reference, currency, urgency) and lines. Spreadsheets are read through column detection, ERP-generated PDFs through their line block. Free-text emails are read conservatively: without a quantity or a number, a line becomes an open question, not a guess.
- Quantity normalisation — units, packs, pack sizes, minimum order quantities and German versus English number formats.
- Matching — candidates from your own catalog, weighed by evidence: your article number, the customer's learned part number, the manufacturer's part number, purchase history, type codes and technical attributes.
- Pricing — a proposal with its source: contract price, customer net price, discount groups, list price or calculation from cost, with warnings such as an expired condition or a margin below target.
- Review and approval — a person checks what is not clear; a second authorised person approves the quote. Nothing is sent automatically.
Worked example: one RFQ, five lines
Fictional example. The request, the customer and the article data are invented to show how the documented rules treat each line; they are not a measurement of match rates.
A maintenance buyer emails a short list. Quotrail recognises the customer from the sender domain and reads five lines. This is how each one ends up in front of the reviewer:
Scroll horizontally to see all columns.
| Customer's line | Evidence found | Status | Reviewer's job |
|---|---|---|---|
| Your article number, 20 pcs | Exact match on your own article number; no contradicting attribute; the next candidate is far behind | Clear | Glance at price and quantity |
| Customer's own part number, confirmed twice before | Learned customer mapping from earlier approved quotes | Clear | Glance at price and quantity |
| Manufacturer part number with one character different | Near match on the number only — weak evidence on its own | Needs review | Pick from the ranked candidates or ask the customer |
| “3 boxes” of an item you sell per piece | Article found; quantity converted with the pack size, the calculation is shown | Clear or Needs review, depending on the match | Confirm the converted quantity |
| “The usual seals, same as last time” | No quantity, no number | Open question | Send the clarification; no line is invented |
The reviewer opens the request and sees the unclear lines first, each with the five best candidates from the catalog and the reasons for their order. When the reviewer resolves a line the customer described with their own part number, that decision is stored in the customer profile — the same spelling matches directly next time.
When the lines are settled, a second authorised person approves the quote. Only then is it a quote: until approval, drafts are marked as drafts, in the PDF too. The approved quote leaves as PDF, Excel, CSV, a CSV file for your ERP import or an email text you send yourself.
What should count as “clear”
A line is only as trustworthy as the rule that let it through. In Quotrail a line is marked Clear only when all four conditions hold:
- there is identifier, customer-alias or purchase-history evidence — similar text alone is not enough,
- no hard technical attribute contradicts the candidate,
- the best candidate is clearly ahead of the second best,
- the article is not discontinued.
If one condition fails, the line stays in review, however similar the description looks. A number that appears for several articles counts for less, so a colliding number cannot create false certainty.
How to evaluate an RFQ tool with your own requests
Vendor demos use clean examples. Your decision should rest on your own requests, including the ugly ones. A short, structured test tells you more than any feature list:
- Collect recent real requests across formats — spreadsheets, ERP PDFs, free-text emails — including the ones your team dreads.
- Import the catalog, customers and conditions you actually quote from, as an export from your ERP.
- For every line, record the tool's status and whether your reviewer changed the result.
- Pay most attention to lines the tool marked clear that your reviewer had to correct. That number decides whether you can trust the fast path.
- Check the exit: can the approved quote reach your ERP in a form your team can import without retyping?
Quotrail works without a live ERP integration: master data comes in as Excel or CSV exports, approved quotes go back as files. A direct integration is scoped separately; it is not assumed.
See it on your own requests
Quotrail is built for technical and industrial distributors whose inside sales teams quote from their own catalog. See the plans and prices, book a call to walk through it with your data, or read how the rules handle units, packs and part number errors.