RFQ Automation for Distributors: Speed Is the Easy Half
An RFQ lands in sales@ at 09:14: a customer's own product names, quantities in a mix of units, one line reading "same as last time". It sits for two days because the rep who knows that account is on holiday, then gets quoted at list price because nobody checks the contract, and you lose it to a competitor who answered on Tuesday.
Fix the two days and you have fixed the obvious problem. You have not fixed the list price, and you have not asked whether that RFQ was ever winnable.
What every vendor in this category sells
Search RFQ automation and page one is unanimous: respond faster. Go Autonomous, Proton.ai, SETVI, ChannelFlex, the CPQ vendors — all speed, and the numbers arrive in a heap:
| Claim | Source shape |
|---|---|
| 60–70% faster (15–20 hrs → 8–12 hrs) | Vendor blog |
| 24–48 hrs → under 15 minutes | Vendor blog |
| 3× faster than manual | Vendor blog |
| 45 seconds vs 15 minutes per RFQ | Vendor blog |
| 3–5 business days is typical | Industry survey |
Notice they disagree with each other. One says the destination is 8–12 hours, another says 15 minutes, a third says 45 seconds. They cannot all describe the same process. What they have in common is that speed is the only axis being measured, because speed is the thing the software obviously does.
The 3–5 day industry baseline is the useful figure in that table, and it sets the real bar: if typical turnaround is three days, consistent same-day response already wins. The marginal value of going from four hours to four minutes is close to zero, because the buyer's decision loop does not run that fast. You are not competing against instant. You are competing against Thursday.
One more number to handle carefully. The "respond within an hour and you are seven times more likely to qualify" research is real, and it is about inbound sales leads — strangers who filled in a form. An RFQ from an account that has bought from you for six years is not that. Applying lead-response urgency to reorder quoting is how teams end up sprinting at the wrong stage of the funnel. If you genuinely have cold inbound enquiries, that is lead qualification and it is a different job.
The bottleneck is intake, not pricing
Ask an estimator what takes the time and they will not say the arithmetic. Once the inputs are known, pricing takes minutes.
The hours go into getting to that point:
- Reading the request and working out what was actually asked for
- Translating the customer's product names into your SKUs
- Checking specifications, substitutions and pack sizes
- Confirming availability and current cost, sometimes with a supplier
- Re-keying all of it into a quoting tool or spreadsheet
Every item on that list is an extraction and lookup problem, which is precisely why this automates well and precisely which part to automate. It is the same machine as order entry, pointed at a request rather than a purchase order: read whatever arrived, resolve it against the catalogue and that customer's history, and produce a structured basket.
Catalogue resolution is the difficult and valuable bit, exactly as it is on the order side. "Same as last time" is answerable if you have the account's order history. "El queso bueno" resolves to a SKU if the mapping has been learned from corrections. A request for an item you no longer stock should surface the substitution, not a blank.
Two decisions that do not automate
Here is where this article parts company with the category.
1. Whether to quote at all
Vendor copy says respond to every inbound RFQ in minutes. In distribution that is not obviously correct.
A share of inbound RFQs — and every experienced sales manager can estimate theirs — are not opportunities:
- Price discovery. A buyer benchmarking their incumbent, with no intention of moving.
- Third-quote padding. Procurement needs three quotes to sign off a decision already made. You are the third.
- Unwinnable specification. The spec is written around a competitor's product.
- Accounts you should not extend credit to, which is a question worth asking before the estimator spends an hour, not after.
Quoting these instantly costs real estimator time, and it does something worse: it teaches buyers that you are the cheap, fast price check. The right automation here is not a filter that silently bins requests. It is scoring the request with the context attached — this account's quote-to-order ratio over the last two years, their last order date, whether the spec matches something you actually win — and putting the marginal ones in front of a person with that history visible.
The metric this protects is quote-to-order ratio. Automate volume without triage and it falls, which is the outcome that gets the whole project cancelled at review.
2. The price
Quoting is where margin is decided, and margin decisions are commercial judgment.
The failure mode is specific and expensive: a system auto-quotes from a stale contract price, or a list price that ignores the customer's tier, or a landed cost that moved when the supplier's acknowledgement changed it and nobody updated the record. You have now sent a customer a number that looks binding. Honouring it costs margin; retracting it costs trust. Speed multiplies whatever your pricing logic does, mistakes included.
So the sane boundary, and the same human-in-the-loop line that applies anywhere money is committed:
- Automate freely where the price is deterministic: contracted price, published price list, standard tier for a known account, no cost movement since the last purchase.
- Require a person for anything with a discount, a new customer, a cost that has moved, a competitive situation the rep knows about, or a value above a threshold you set.
The useful thing an agent does in the second case is not decide — it is assemble. The rep opens one screen holding the request, the resolved basket, this customer's last three quotes and what they converted at, current cost, margin at each price point, and stock position. The decision then takes ninety seconds instead of sitting for two days waiting for someone to gather all that.
The quote nobody follows up
The last gap in the category: everything stops at sent.
A quote is not revenue. It is an offer with an expiry date, and in most distributors a large share of them are never mentioned again. Nobody chases, the validity period lapses quietly, and the effort that went into producing it is written off without ever being counted.
This is the same machinery as receivables follow-up, pointed earlier in the cycle:
- A short follow-up on a schedule you set, referencing the quote number and what it covers
- A different message when the validity date is approaching, because that one has a legitimate reason to exist
- Stop the sequence the moment the customer replies or orders
- Record the outcome, including losses and why, so quote-to-order ratio becomes a real number rather than a guess
That last point compounds. Once you record why quotes are lost, you find out whether you are losing on price, on lead time, or on availability — and those three have completely different fixes. Most distributors assume price and are wrong at least some of the time.
Measure four things
- Quote-to-order ratio, by customer. The number that tells you whether faster quoting is winning business or just producing more quotes. Watch it fall if you automate volume without triage.
- Same-day response rate. Not average turnaround, which one terrible outlier ruins. The share answered the day they arrived, against an industry baseline of three to five days.
- Share of quotes priced without human input. The direct efficiency number, and it should stay well below 100% forever. If everything is auto-priced you have automated your margin decisions.
- Quotes expiring with no follow-up. Usually the biggest number on this list the first time anyone measures it, and the cheapest to fix.
Deliberately absent: quotes produced per day. It rewards exactly the behaviour that destroys the first metric.
Start here
Pull the last hundred RFQs. For each, record how long it took to answer, whether it converted, and if not, why.
Two findings are close to universal. A meaningful group came from accounts that have never converted a quote — those did not need to be faster, they needed to be triaged. And a group was never followed up at all, which means the quoting effort was spent and then abandoned.
Neither of those is a speed problem, and both are larger than the speed problem. Fix the response time too, by all means; it is the easy half and the software does it well. Just do not let it be the only thing you fix, because the vendors selling this category have no incentive to mention the other half.
Oido runs quoting end to end for distributors and manufacturers: reading RFQs out of the shared inbox, WhatsApp or PDF attachments the way it reads incoming orders, resolving the customer's own product names against your catalogue and history, pulling contract pricing and stock from your ERP, auto-quoting only what is deterministic and assembling everything a rep needs to decide the rest, then following the quote up until it converts or is honestly recorded as lost. Heaviest fits are wholesale trade and manufacturing. Send us a hundred RFQs and we will tell you how many were worth quoting.
Frequently asked questions
What is RFQ automation?
Software that takes an incoming request for quotation, reads the items and quantities out of whatever form it arrived in, resolves them against your catalogue, retrieves the right price for that customer, and produces a quote without a person retyping anything. The good implementations stop short of sending it unpriced-decisions-and-all, because quoting is where margin gets decided.
How fast should you respond to an RFQ?
Faster than the competitor who is also quoting it, which in most of distribution means same day rather than same minute. Industry surveys put typical manufacturing quote turnaround at three to five business days, so consistent 24-hour response already differentiates you. Chasing sub-minute response is optimising a variable that stopped mattering once you were first.
Should you quote every RFQ you receive?
No, and this is the part vendor guides skip. A meaningful share of inbound RFQs are price discovery from buyers with no intent to switch, or third-quote padding to validate a decision already made. Quoting all of them instantly costs estimator time, trains buyers to use you as a price check, and depresses your win rate metric until nobody trusts it.
What is the risk of automated quoting?
A fast wrong price. Auto-quoting from a stale contract price, a list price that ignores the customer's tier, or a cost that has moved since the last purchase produces a binding-looking number you have to honour or retract, and both are expensive. Speed multiplies whatever your pricing logic does, including its mistakes.
How is an RFQ different from a sales lead?
A lead is someone who might buy something. An RFQ is usually an existing or known customer asking for a price on a specific basket, and it is a step in a buying process already underway. Lead-response research about answering within minutes is about the first kind and gets misapplied to the second, which is how teams end up optimising the wrong stage.
What actually slows down quoting?
Not the pricing maths, which takes minutes once the inputs are known. It is the intake: reading the request, interpreting a customer's own product names, checking specifications, confirming availability and cost, and re-keying it all into a quoting tool. That is an extraction problem, which is why it automates well.
Do you need CPQ software for RFQ automation?
Only if your products are genuinely configurable with interdependent options. Most wholesale and distribution quoting is catalogue items at customer-specific pricing, where CPQ is heavy machinery for a job your ERP already holds the data for. Start by automating the intake and the lookup, and buy configuration software when configuration is the actual bottleneck.