How to write an RFQ: a worked example, line by line

An RFQ has one job: make five different companies answer exactly the same question, so that comparing their answers is arithmetic instead of interpretation. Every part of writing one serves that job. The moment two suppliers can read your request differently, you no longer have five comparable quotes; you have five small essays about what each supplier hoped you meant.

The good news: an RFQ that does its job is not longer than one that fails. It is just more decided. Here is the structure, a worked example, and the specific places where ambiguity sneaks in.

If you would rather start from a file, there is a free, editable RFQ template in English and Arabic with all of this built in.

The five sections every RFQ needs

1. Request details. Reference number, issue date, quote deadline (with a time and a timezone), how long quotes must stay valid, the currency, the delivery location, and the required delivery date. This block takes ten minutes and prevents the two most common quote-comparison failures: prices in mixed currencies and offers that expire before you decide.

2. Line items. The heart of the document. One line per priced thing, each carrying a description, a specification or standard, a quantity, and a unit. This is where comparability is won or lost; more on it below.

3. Commercial terms. Payment terms, delivery terms (who pays freight, insurance, customs), warranty expectations, penalties if any. State what you expect; let suppliers flag what they cannot meet. Terms discovered after award are renegotiations you funded.

4. What the quote must include. Price per line and total, delivery confirmation, explicit exclusions, and the company documents you need (trade license, relevant certifications). Asking for exclusions explicitly is the cheapest risk control in procurement: it converts silent scope gaps into written statements you can compare.

5. Evaluation. How you will decide, stated plainly. Even one sentence, such as "total cost 60%, delivery 20%, warranty and compliance 20%", changes what suppliers optimize for. Add the clarification rule here: questions to one named contact by one date, answers shared with all bidders.

A worked example

Say an operations team needs office chairs and sit-stand desks for a new floor. Here is the difference between a line item that produces comparable quotes and one that produces email:

Weak: "Office chairs — 40 pcs."

Forty of what? Task chairs or executive chairs? Any standard? Warranty? Each supplier resolves the ambiguity in their own favor: one quotes a basic chair to win on price, another quotes mid-range and looks expensive, and the comparison is fiction.

Strong: "Ergonomic task chair — adjustable seat height 42–53 cm, lumbar support, armrests, castors for hard floor, upholstery fabric, load rated ≥110 kg, EN 1335 or equivalent — 40 — pcs."

Now every quote answers the same question. The specification does not need to be exhaustive; it needs to close the interpretations that would change the price. "EN 1335 or equivalent" does more work than three paragraphs of prose, because it imports a whole standard by reference.

The full request would carry perhaps six such lines (chairs, desks, monitor arms, delivery and installation as its own line), the commercial block ("delivery to Floor 9, Building X, by 30 September; payment 30 days from delivery; 2-year warranty minimum"), and the evaluation sentence. Two pages. That is a complete, professional RFQ.

Where RFQs go wrong

Lump sums where lines should be. "Supply and install office furniture — 1 lot" is unpriceable except as a guess with margin on top. Every lump sum in your request is a place where the low bidder and the good bidder become indistinguishable. Itemize; if the request is quantity-heavy, use a bill of quantities and let the structure carry the work.

Specifying the brand instead of the requirement. Naming one manufacturer's model as the specification either locks you into that supplier's pricing or forces everyone else to guess what "equivalent" means. Specify the performance and the standard; accept equivalents explicitly.

Vague deadlines. "Please respond soon" produces a slow trickle of quotes you cannot compare because the early ones expired. One deadline, one validity period, in writing.

Private clarifications. When one supplier asks a good question and you answer only them, you have quietly forked your RFQ: bidders are now pricing different documents. Every answer goes to every bidder, always. This single habit removes a whole class of award disputes.

Accepting substitutions inside lines. A supplier who swaps your specified item for "similar" inside a line has broken the comparison for everyone. Take alternatives (they are often genuinely valuable) but as clearly-marked separate notes, priced separately.

The step most guides skip: what happens to the quotes

Writing the RFQ is half the discipline. The other half is holding the structure when responses arrive, because they will arrive as PDFs, spreadsheets in personal formats, and prices buried in cover letters, and the first thing most teams do is build a comparison spreadsheet by hand and introduce transcription errors into their own decision.

This is the part that structure fixes outright. On VEXORS the RFQ is the structure: you publish the line items, invite your suppliers or discover new ones by category, and suppliers price each line directly, free for them. Quotes arrive already aligned, item by item; clarifications run in one thread every bidder sees; and each bid is scored by AI against your request with the reasoning written out. The comparison you would have built by hand at midnight simply exists when bidding closes, with a timestamped record behind the award.

A pre-send checklist

Thirty seconds before you send, check:

If all seven hold, you have written an RFQ that will do its one job: come back as answers you can put side by side and decide on.

The fastest way to see the difference is to run one. Create a free account, build the request from your line items (typed, imported from a spreadsheet, or drafted by AI from a plain-language brief) and watch comparable quotes arrive.