The Unanswered RFQ Is the Most Expensive Thing in Your Inbox

It is a little after eight on a Tuesday and the shared inbox has 140 new requests in it. Most came in overnight through NetComponents and Brokerforum, a few through IC Source, a handful direct from buyers who have your address. Someone on the sales desk opens them one at a time. The ones you have on the shelf get a quote. The ones you can source by Thursday get a phone call. The rest get read, maybe skimmed, and deleted.

By ten the inbox is clean. Nobody wrote down what was in the deleted pile.

That pile is what this piece is about. I think it is the most expensive thing in the building. Dead stock in the back at least shows up on a report once a year. The requests you could not answer show up nowhere, and they were the market telling you, in plain text, what it wanted to buy from you.

I should say what I am and am not before going further. I have never worked a stockroom or quoted a part across a counter. I ran a digital agency for almost 18 years, and now I use AI to fix specific problems inside businesses. What I know is how work moves through an office and where it quietly piles up. A distributor’s request stream is one of the clearest piles I have seen. I designed a demand ranking and a market hunt for a parts distributor whose stream had outrun what any person could read, and the same method fits any distributor that buys on judgment and sells through a marketplace feed.

What an unanswered request actually tells you

An RFQ you could not quote is not noise. It is a buyer handing you four facts in one line: the part number, the quantity, how soon they need it, and, if you look at the sender, who they are. You did not pay for any of it.

One request on its own says very little. Forty requests for the same part from twelve different companies across three months say something else entirely. They say there is a part the market wants, that you do not carry, and that the buyers who want it keep coming to you anyway.

Nobody counts them because the inbox is read for a different job. The sales desk reads it to answer the question “what can I quote today.” That is the right question for the desk. It is the wrong question for the buyer, and both of them are working from the same inbox.

Three columns worth keeping

If I could keep only three numbers per part number, these are the three:

• Buyers: how many distinct companies asked for the part in the window.
• Requests: how many times it was asked for in total.
• Urgent share: what fraction of those requests carried a near deadline, a “need by Friday”, a line about a stopped production line.

Buyers matters more than requests. One purchasing agent blasting the same RFQ to every distributor on the marketplace every week inflates the request count without telling you anything about demand. Twelve companies asking once each is a market. One company asking twelve times is a person with a problem, and quite possibly a person who already solved it somewhere else.

Urgent share tells you something about price and something about the buyer. A part that is always requested with a deadline attached is a part where the buyer has run out of easier options. That is a different negotiation from a part people are pricing for a build next year.

Count those three by part over 90 days, then set the list against what is on the shelf. The top of the list, filtered to parts you do not stock, is what the market has asked you for and you have said no to.

The same ranking tells you what to reprice and what to stop buying

Put the ranking next to the stock list and three groups fall out on their own.

• Ranked high, not stocked. This is the acquire list. It is the one everybody expects.
• Ranked high, stocked, still on the shelf. Something is off. Either the price is above the market, or quotes on this part go out too slowly, or the trace grade you hold is not the grade the buyers want.
• Stocked, bought on a hunch, nobody has asked in two quarters. This is the stop-buying list, and probably the reprice-down list.

That third group is the uncomfortable one, because it is usually someone’s call from a year ago. A last-time-buy notice came through, the part looked like it would be scarce, a pallet was bought. The ranking does not say the call was wrong. It says the market has not shown up for it yet, which is a fact worth knowing before the next hunch.

Picture a Monday

This scene is made up.

Before. Monday morning, the inbox has the weekend’s requests in it. The sales lead and the buyer are both reading it. The sales lead quotes what she can and deletes the rest. The buyer scrolls past the same requests looking for anything that rings a bell, remembers a part he saw asked for last month, cannot recall who asked or how many, and goes back to a broker stock offer sheet that landed in the same inbox. By the afternoon he has bought two lines from the sheet because the price looked right. Nobody asked whether anyone wanted them.

After. Same Monday. The buyer opens a one-page list, 25 part numbers, three columns each: buyers, requests, urgent share. It covers the last 90 days of marketplace exports and the inbox. Parts already on the shelf are marked. The top unstocked line shows nine buyers, 31 requests, a third of them urgent. He spends the morning on that one part: what the alternates are, who has it, at what grade. The broker offer sheet gets read against the list instead of on its own, and the two lines that looked cheap are not on it. He passes.

Nothing in the second version required a new system. It required the deleted pile to be counted.

What has to be in place first

• An export of request history from each marketplace you quote from. NetComponents, Brokerforum and IC Source each hold your RFQ history in some form. The layouts differ. The fields you need are the same: part number, quantity, date, sender.
• The shared inbox, exported or forwarded. Direct requests from buyers who email you are often the strongest signal, because those buyers chose you rather than the marketplace.
• A stock list from the ERP: part, quantity on hand, cost, date received. It does not have to be clean. It has to be current.
• A way to treat part numbers that buyers typed five different ways as the same part. Suffixes, packaging codes and spacing all vary.
• One person who owns the list and looks at it every week. A list nobody owns is a spreadsheet.

Guardrails worth keeping

• The ranking is a question, not an order. A part at the top of the list with no supply at any grade is still a part you cannot sell.
• Count requests, not your own quotes. Quotes measure what you did. Requests measure what the market wanted.
• Keep buyer identities inside the building. The list is built from people who trusted you with a request.
• Watch for one sender doing the work of twelve. The buyers column exists for this reason.

How you’ll know it’s working

• Quote rate: the share of incoming requests you could answer, month over month. This is the headline number.
• Coverage of the top 25: how many of the 25 highest-ranked parts you now stock or can source on a known lead time.
• Turns on parts added from the list, compared with turns on parts bought the old way.
• Response time on top-25 parts.
• Dead stock as a share of shelf value. If the stop-buying list is doing its job, this falls slowly and keeps falling.

Where I’d start

One quarter of history and the top 25 parts. Pull 90 days from each marketplace and the inbox, count buyers and requests by part, mark what is stocked, and read the top 25 unstocked lines with the buyer in the room. If that hour does not change at least one purchasing decision, stop there. If it does, do it again next month.

The free week on the Hunt page builds this list from your own exports as its first step.