The quote goes out in minutes, not hours
A customer sends a messy list — an email, a spreadsheet, a photo of handwriting. The system reads it, drafts the quote with your margins applied, and a person checks and hits send. If quoting speed decides your win rate, this is the highest-leverage software you can own.
- 1
- workflow in phase one — the quoting engine
- 3
- phases, each justified by the last one's measured result
- 0
- ERP replacements involved
Who this is for
If this sounds like your company, keep reading
You sell sourcing, not shelf stock. Customers send you their lists — industrial supplies, parts, safety gear, anything — and your product is the searching, quoting and chasing itself. Today that runs on email, spreadsheets and somebody's memory.
A strong fit
- A distribution or sourcing business where requests arrive as email, attachments, photos and phone calls
- High volumes of small orders, where margin depends on how cheaply each one is processed
- Quoting that takes hours because a person re-types, deciphers and hunts prices by hand
- Supplier knowledge that lives in one or two veterans' heads
- Runs the business on email, spreadsheets and QuickBooks — and would rather keep it than replace it
Not what we would sell you
- A full ERP rebuild — you almost certainly do not need one
- Demand-forecasting AI when you hold no inventory — anyone selling you that did not study your business
- A year-long platform build — phase one has to be one workflow that pays for itself
- Software that removes your people from the quote — judgement and relationships are your actual moat
The problem
The list arrives, and the clock starts
A customer emails a list of items. Sometimes a spreadsheet. Sometimes a photo of handwriting. Sometimes “we need those valves like last time.”
Then someone on your team re-types every line, works out what “those valves” actually means, hunts prices from suppliers, builds the quote and sends it — hours later. Meanwhile the customer sent the same list to your competitors, and the first credible quote back usually wins.
The re-typing is not the hard part. The hard part is turning a customer's own words into specific items at specific prices, reliably enough that a person does not have to re-check every line. That is what the engine is for.
Your team does not disappear. The typing does.
The system reads the list, drafts the quote, and marks the handful of lines that need a person's judgement — so your people spend their time on exactly those, and on the relationships that win the next order. A person presses send on every quote, always.
The complete workflow
One request, start to finish
This is what happens to a single customer request once the system is in place — the whole journey, with a person in exactly two places: reviewing the marked lines, and pressing send.
- 1
Every request lands in one place
Email, a spreadsheet, a scanned document, a photo of handwriting, a forwarded thread. One queue, one answer to “what have we not replied to yet”, and every request shows how long it has been waiting.
- 2
The system knows who is asking
Which unlocks their pricing, their history, and the names they use for things. When it cannot tell, it says so and holds the request for a person rather than guessing.
- 3
The message is split in two
The requested lines on one side — each with its quantity. The deadline, the job reference, the correction and the question on the other. Those ride on the quote, because a deadline nobody carried is an order lost late.
- 4
Every line becomes a real item
Messy descriptions, superseded part numbers and “like last time” are matched to specific items — by part number, by what this customer bought before, by description. Each line carries its own confidence.
- 5
Suppliers are found and compared
For each line: who can supply it, at what cost, on what lead time — side by side, with your margin computed at the price you are quoting. The decision that sets what the order earns, made instead of skipped.
- 6
A person reviews the marked lines only
The uncertain lines are flagged; the rest are already right. Judgement is spent only where judgement is needed — that is the entire saving.
- 7
A person presses send
The quote goes back on the channel it arrived on, in the same thread, with the deadline and the customer's instructions carried on it. You are first back while competitors are still typing.
- 8
What was learned is kept
Confirmed matches — “those valves” means this exact item, for this customer — are recorded, so next time that line needs nobody. The share of lines needing review falls the longer it runs.
What we build
One backbone, three phases, each one earning the next
We do not propose a platform. We propose a first phase small enough to be provably right — and every later phase is a separate conversation, had with the previous phase's numbers on the table.
- 1
The quoting engine — internal only
The land: smallest scope, fastest payback, zero customer-facing risk.
Requests arrive from email, a spreadsheet, a scanned document or plain prose. The system pulls out every line with its quantity, makes sense of messy part descriptions, drafts the quote with your margin rules applied, and marks the uncertain lines. Your team reviews and sends — from hours to minutes on the requests you already handle today.
- 2
The customer portal — the front door on the same engine
Many distributors' websites already promise a portal; this makes it real.
A real login for your registered customers: submit requests, upload lists, see quote status, approve, attach purchase orders, track deliveries. Nothing new underneath — it is the phase-one engine with a door on it, and the moment it ships, your website's promise is true.
- 3
Intelligence and assistants — each one its own decision
Built only where phase two's numbers say they will pay.
Supplier matching that turns your veterans' memory into something anyone can search, so new hires quote like twenty-year hands. Chase emails that send themselves, read the replies and update every order's status before the customer has to ask. Supplier offers parsed into a side-by-side comparison. Per-customer spend reporting that makes your contract renewable on evidence. And the quiet back-office glue — purchase order to receipt to invoice — synced to the accounting you already run.
Said before you ask
What we would tell you not to buy
You probably do not need an ERP. If you hold no inventory, you do not need demand forecasting. And if a vendor's first proposal is a year-long platform build, walk away — including from us.
The work worth doing sits in three places: the re-typing, the searching, and the chasing. Those become software. Judgement and relationships stay with your people, because that is the part of your business a competitor cannot copy. We would rather tell you what will not work and earn the second meeting.
The same workflow, as a product
Where this thinking comes from
This page is the engagement version of a workflow we also build as a product. Quote Rabbit is the place quoting happens for distributors on Microsoft Dynamics 365 Business Central — it connects to their own system and works against the catalogue, customers and order history live.
If you run Business Central online, start with Quote Rabbit — it is the faster path to the same outcome. If you run anything else, this engagement is how the same workflow gets built on your stack, phase by phase.
Common questions
Asked by owners, answered honestly
Including the two questions that matter most: whether you need an ERP at all, and whether we can show you a reference — where the honest answer is not yet.
Do we need an ERP for this?
Probably not, and we will say so on the first call. What most sourcing businesses need is an order-management backbone that talks to the accounting they already run — usually QuickBooks, which stays exactly where it is. A full ERP project is the thing this engagement is designed to make unnecessary.
Should we go through an agency, hire a freelancer, or find an employee?
The honest version: a freelancer is one person carrying your whole system — vacation, illness or a better offer stalls the project. An employee is a large yearly cost, and one developer cannot cover design, the AI work and the infrastructure anyway. An agency gives you a team and accountability, but only makes sense if it scopes small. The real trick is not who you hire — it is starting with a build small enough that you can afford to be wrong about it.
What if the AI reads a line wrong?
A person presses send on every quote, always. Each line carries its own confidence, so the uncertain lines are marked for review and the rest are not — and a line the system cannot make sense of says so plainly, with the customer's own words next to it, rather than guessing quietly.
What exactly does phase one include?
The quoting engine, used internally by your own team: requests come in from email or a file, the lines are read and matched, the quote is drafted with your margin rules applied, and your people review and send. Nothing customer-facing yet, which keeps the risk low — and its measured result is what justifies phase two, or tells us to stop.
We run Microsoft Dynamics 365 Business Central. Is this the same as Quote Rabbit?
Same workflow, different shape. Quote Rabbit is our product version: it connects to your own Business Central and works against your catalogue, customers and order history live. If you run Business Central online, start there — this page is for businesses on other systems, where we build the same workflow as an engagement.
Can you show us another distributor using it?
Not a named one, and we will not invent a reference. This page comes out of real conversations with distributors, and we run production AI systems for clients in other industries — but in this niche, what we offer first is a working session on your own lists: send a real request from your inbox and watch it become a reviewed quote.