---
title: "AI Automation for Trading Companies | JustAutomate"
description: "Wholesale and B2B distribution without re-keying: a draft quote from an emailed enquiry, orders entered with price control. Workshop EUR 3,500 net per week."
url: "https://justautomate.ai/en/industries/trading-companies/"
lang: en
---

Wholesale and B2B distribution

# AI automation for trading companies from enquiry to order

In wholesale the sale runs through a salesperson. The enquiry arrives by email or by phone, the price list is negotiated per customer, and someone still re-keys the order into the system. We take that second part off the salesperson: a draft quote is built from current prices and stock, the order enters your ERP with every line checked, and the conversation with the customer stays with a human.

 Book a free consultation [See how the workshop works](https://justautomate.ai/en/services/process-workshop/)

- 14 daysto the first working automation, counted from the end of the workshop
- EUR 3,500net per week of workshop, a published price we do not negotiate
- 5 to 10 secondshow long the AI agent takes to answer what price this customer gets, what is in stock and when the next delivery lands, from data in your own systems

The problem

## Why does a quote for a regular customer take half a day?

The enquiry arrives as one sentence: "price me 40 units of the item we bought last time, when can you ship". The salesperson opens the price list, looks up the customer's discount group in the ERP, checks stock, calls the supplier for a lead time and builds the document in a spreadsheet. An hour, and on a larger basket half a day. In that time the customer has sent the same enquiry to two other distributors.

Then the order comes back by email, often as "confirmed, but swap item A for item B", or as a table pasted out of the customer's own system. Someone re-keys it line by line. A wrong item code or a wrong price does not surface during entry, it surfaces on the delivery note or on the invoice, when the goods are already on a truck. **The salesperson sits in admin instead of on the phone**, and it is the calls that close the sale.

The cost of that delay is countable, because wholesale sells on trade credit. [Directive 2011/7/EU on late payment in commercial transactions](https://eur-lex.europa.eu/eli/dir/2011/7/oj) sets the frame: where the contract fixes no date, payment falls due 30 days after the invoice, a business to business contract should not push the term past 60 days, and a creditor paid late is owed at least EUR 40 per invoice towards recovery costs. So two months routinely pass between sending a quote and seeing the money, and the owner learns the margin on the deal only after the month is closed.

- The enquiry sits in the inbox until the evening, because the salesperson is out at a customer
- The quote goes out with a price from an old list, or with another customer's discount
- An order re-keyed from an email carries the wrong item code, and it shows on the delivery note
- The customer calls to confirm a shipping date nobody ever sent him
- The margin on the deal is known once the month is closed, which is after the fact

Why now

## What changed that a wholesaler can wire this up in weeks?

Email stopped being a wall. A language model reads an enquiry written as a plain sentence, a pasted table and an attachment, and pulls item codes, quantities and the expected date out of all three. That is the part that blocked automation in wholesale for years: enquiries never arrive in one format, because every customer writes the way he writes.

The second thing is where trading companies keep their data. **A large share of wholesalers still run their ERP on a server in their own office**, with the database in the building rather than in the cloud, so nothing on the internet can query it directly. We reach it with a service running inside your network that reads the database and exposes only the fields we agree on, or through the vendor's own integration layer where the system has one. That is not an obstacle, it is a different way of connecting.

The third change is the AI agent. A rule will copy lines from an email into the ERP, but it will not answer a salesperson asking "what price did we give this company in March and how much credit limit is left". BOND, our company AI agent, answers that in a sentence in 5 to 10 seconds, from the price list, the customer history and the agreed trading terms.

What changes

## What do the sales desk and the order desk get?

- ### The draft quote is waiting before you call back
The emailed enquiry is broken down into lines, customer and requested date, and the system builds a draft quote out of that. The salesperson checks it, corrects it and sends it under his own name.
The salesperson approves a finished document instead of building one in a spreadsheet
- ### The price always comes from the live list and the customer record
Discount, price group and individually agreed terms come straight out of the customer record, not out of somebody's memory. A supplier price change lands in the quote the same day.
Terms read from the customer record in your ERP, not from a file saved last quarter
- ### An emailed order enters the system with checks
Lines, quantities and prices are matched against the product records before the document is saved. A mismatch stops the order and goes back to the salesperson with a note on what does not add up.
A price or item code mismatch shows before the delivery note, not on the invoice
- ### Availability and lead time confirm themselves
Once the order is accepted the customer gets a confirmation with the lines, the stock position and the planned shipping date. Items that are short are listed separately, with the date from the supplier.
The customer gets a confirmation instead of phoning the salesperson for one
- ### Quotes with no reply chase themselves
A quote with no reaction comes back to the salesperson as a task, with the line items and the date it went out. It does not get buried just because that week had more urgent things in it.
A reminder after 3 and after 7 days, with the conversation history on the CRM deal
- ### Margin per line on the day of the deal
Purchase price, discount, freight and packing cost set against the selling price for this customer. You see it on the document rather than after the month is closed.
Margin counted per line and per customer from documents you already have

Proof

## Do you work this way yourselves?

Yes, and we will say up front what the number does NOT prove. This is not a wholesaler, it is our founders' own agency, so we are not showing you the result of a trading company. We are showing what happens to the cost of delivery in a 30-person business when an AI agent takes over the admin around sales. Selling in wholesale depends on the salesperson and we are not changing that.

Built for ourselves. These are not client deployments but implementations inside the founders' own companies. We say so plainly, because proof you cannot verify is not proof.

- ![JustIdea Agency](https://justautomate.ai/klienci/justidea.png)![JustBot](https://justautomate.ai/agenci/justbot-80.webp)JustBottheir agent
JustIdea Agency, a performance marketing agency of around 30 people
23
people from the company work with the agent
38
systems wired in permanently
112
processes run without a human
475
thousand lines of the agent’s code
An agent wired into Google Ads, Meta, the CRM, the task system, the SEO tool, Analytics, Search Console and the data warehouse. First commit 1 March 2026, today 23 people use it.
8 of 38 systems it works in

  - Google Ads
  - Meta Ads
  - HubSpot
  - ActiveCollab
  - Ahrefs
  - GA4
  - Search Console
  - BigQuery
Before
In the twelve months before launch the agency ran at a margin of minus 4.42 percent. Across a twenty-month window there was not a single run of three profitable months.
After
May, June and July 2026 are three profitable months in a row, the first such run in the whole window. Revenue barely moved, so the change sits on the cost of delivery, not on sales.
Margin up 5.55 percentage points year on year, with revenue up 1.5 percent

What about us

## What if trading works differently here?

Four sentences come up in wholesalers more often than any others. All four are fair, and we have a concrete answer to each.

- Our customers like to call and talk, we do not want to take that away.
Nobody is taking it away. We automate what happens AFTER the call: the note, the quote, entering the order, confirming the date. A salesperson who is not re-keying line items has more time for the phone, not less.
- Our ERP sits on a server in the office, it does not connect to anything.
It connects, just differently from a cloud application. The database is inside your network, so we put a service there that reads it, or we use the vendor's integration layer where one exists, and it exposes only the fields we agreed. The scope of access is written down at the workshop before anything is plugged in.
- Every customer has his own price list and his own discounts, no rule will handle that.
It does not have to, because those terms already sit in the customer record and in the price lists. The system reads them instead of inventing them. Exceptions that were never written down stay a decision for the salesperson, and we leave it that way: he knows them, and we are not going to pretend they can be guessed.
- We are afraid the machine will send a customer the wrong price.
That is why the quote goes out only after the salesperson approves it. The system builds a draft and shows where each price came from, and sending is a separate click. On incoming orders it works the other way round: a price or item code mismatch stops the document rather than letting it through.

CRM

HubSpot, Pipedrive, Salesforce, ActiveCampaign

ERP, stock and sales systems

Sage, Microsoft Dynamics 365 Business Central, SAP Business One, on-premise systems read through their database

Invoicing and accounting

Invoicing tools over their API, accounting exports, document workflow

Online sales and marketplaces

Shopify, WooCommerce, PrestaShop, BaseLinker

Mail, spreadsheets and data

Google Workspace, Microsoft 365, Google Sheets, Excel, BigQuery

Automation layer

n8n, Make, custom integrations over an API

## The trading cycle step by step: who does it today, what the automation takes over

The path from enquiry to invoice in a typical B2B wholesaler. The third column is the scope that can realistically come off the salesperson without taking the customer conversation away from him.

| Step                         | How it works today                                    | What the automation takes over                                                        |
| ---------------------------- | ----------------------------------------------------- | ------------------------------------------------------------------------------------- |
| Emailed enquiry              | the salesperson reads it and lists the items on a pad | item codes, quantities and the requested date pulled from the body and the attachment |
| Price and discount           | the price list plus memory of the customer terms      | the price list and the customer record read from the ERP                              |
| Availability and lead time   | a call to the warehouse or to the supplier            | stock and the next delivery date read from the system                                 |
| Quote                        | a document built in a spreadsheet and saved to PDF    | a document from a template, sent once the salesperson approves it                     |
| Customer order               | re-keyed from the email line by line                  | entered with item codes, quantities and prices checked                                |
| Confirmation to the customer | sent when the salesperson gets round to it            | goes out when the order is accepted, with the shipping date                           |
| Quote with no reply          | the salesperson remembers it, or nobody does          | a reminder after 3 and after 7 days, with the line items                              |

We do not take every row at once. At the workshop we pick the two or three with the highest return and start there, because the first rollout is meant to show an effect, not to cover the whole company.

How we work

## Five steps from the sales inbox to the order in your system

Every step ends with something you can judge before you decide on the next one.

1. ### Workshop
One week. We walk the road from enquiry to invoice together with the salespeople and whoever handles orders. We count how long a quote takes today, how many lines go through manual re-keying and where the price mistakes come from. We come out with a map and a list of the places where the same number is typed in twice.
2. ### Connecting the system
We check on your data whether customer records, price lists, stock and order history can be read safely. On an ERP that lives on your own server this means a service inside your network or the vendor's integration layer, with the field list written down on paper. This step ends with a read, still without any automation.
3. ### First automation in 14 days
In wholesale we usually start with the draft quote from an emailed enquiry, because that is what takes the most hours off the salesperson. It runs alongside the current way of working: the salesperson gets a finished document and decides himself whether to use it. The comparison with what he would have done by hand is visible from day one.
4. ### Orders and confirmations
The second area is order entry with line and price checks, plus automatic confirmation of availability and shipping date. We agree the thresholds at which a document stops for a human decision: a new customer, a price outside the band, an item that is not in the product records.
5. ### Measurement after 30 and 90 days
We go back to the numbers from the workshop: time from enquiry to quote sent, the number of price corrections, the number of orders held by the checks. Only on that basis do we decide whether the next area is worth taking.

Before and after

## How a wholesale salesperson's day changes

### Before

- The morning goes on listing items from three enquiries, before he calls anyone
- He checks stock and lead time by phoning the warehouse, because he does not trust the system
- He re-keys the order from the email into the system, line by line
- He remembers last week's quote at the moment the customer buys elsewhere

### After

- The morning starts with three draft quotes to check and sign off
- Stock and the next delivery are already in the draft quote, with no phone call
- The order is already in the system and he only looks at the lines with a mismatch
- A quote with no reply comes back to him as a task after 3 days

The first draft quotes are visible on the day of launch. Time from enquiry to quote sent and the number of price corrections are judged after 30 days, against the numbers measured before the start.

For the owner and the sales director

## What the person signing this gets out of it

The first thing shows within a week: you stop being the bottleneck on non-standard quotes, because the customer terms are in the system rather than in your head. A salesperson going on holiday stops meaning that his customers wait for a quote until Monday.

The second one matters more in a conversation with a bank or a partner. The margin on a single deal stops being the private knowledge of whoever closed it and becomes a number in a system that you can put on the table and defend. The same goes for trading terms: they are written down, not agreed on a phone call some time ago.

- Customer trading terms written down and versioned, independent of one person
- Time from enquiry to quote measured before the rollout and again after 30 days
- Price mistakes caught before the document is issued, not after the delivery
- The first automation runs alongside the current process, so nothing stops

What it costs

## The workshop price is public, the rollout is quoted after the process map

The process workshop costs **EUR 3,500 net per week** and we publish that number before the first call. It runs from 1 to 7 weeks, and a wholesaler usually needs one or two: the first for the enquiry and quote path, the second for orders and settlement if you sell through several channels at once.

We quote the rollout after the workshop, once we know how many systems we are connecting to and what state the customer and product records are in. Tools and AI models are paid straight to the supplier with no margin from us, and model usage in this kind of workflow runs at EUR 0.005 to 0.07 per query. On your own n8n on your own server you pay nothing for the automation layer itself.

- ### Process workshop
EUR 3,500 net
per week, in wholesale usually one to two
A closed piece of work. It ends with a map of the trading cycle and a 30 day plan you can run without us.

  - A week of work with the salespeople and the order desk
  - A map of the road from enquiry to invoice, with times measured on site
  - An assessment of what can be read out of your ERP or stock system
  - The two or three automations with the highest return, with the hours counted
  - Metrics to measure after 30 and 90 days
 Start with the workshop
- Most popular
### Workshop and first rollout
quoted after the workshop
the scope follows from the number of systems and the state of the data
The variant most companies pick. The draft quote from an emailed enquiry is working 14 days after the workshop ends.

  - Everything in the workshop
  - A connection to your stock and sales system, on-premise included
  - A draft quote from the enquiry, priced from the customer record
  - Order entry with line and price checks
  - Training for the person who runs it on your side
  - Measurement after 30 and 90 days
 Talk to us about a rollout
- ### Trading cycle and AI agent
quoted after the workshop
for teams where questions about price and stock come up all day
Everything in the rollout plus BOND, the company AI agent that answers salespeople from the price list and the customer history.

  - Everything in the first rollout
  - An AI agent answering on customer price, stock, lead time and order history
  - Availability confirmations and reminders on quotes with no reply
  - Alerts on prices outside the band and on credit limits exceeded
  - Access for sales, the order desk and the warehouse
 Ask about the agent

The workshop ends with a document that stands on its own: a process map, a list of places worth automating and the metrics to measure. You can run it with your own IT person or with another supplier. If we conclude after the workshop that there is nothing here worth automating, we say so plainly and we do not propose a rollout.

## Frequently asked questions

What exactly do you automate in a trading company?

Four things most often: building a draft quote from an enquiry sent by email, entering the customer order with item and price checks, sending a confirmation of availability and shipping date, and chasing quotes that got no reply. On top of that sits an AI agent that answers salespeople from the price list and the customer history. The scope of the first rollout is chosen at the workshop, usually two or three items from that list.

Can you connect to an ERP that runs on our own server?

Yes, by a different route than a cloud application. The database sits inside your network, so we put a service there that reads it, or we use the vendor's integration layer where the system has one, and that service exposes only the agreed fields outward. The scope of access is written down at the workshop and nothing goes beyond what we agreed.

Our customers prefer to call. Does this take anything away from them?

No, because we automate what happens after the call, not the call itself. The customer phones exactly as before, and instead of scribbling item codes on a pad the salesperson already has a draft quote built from current prices and stock. The effect is the opposite of the worry: there is more time for the phone, because the admin stops eating it.

Who is responsible for the price on a quote the system built?

The salesperson, and that is deliberate. The system assembles a draft and shows which price list and which customer record every line came from, but sending it is a separate human decision. On incoming orders the rule is reversed: a price outside the agreed band stops the document and sends it back to the salesperson instead of letting it through.

How long does the rollout take and where do you start?

A week of workshop, then 14 days to the first working automation. In wholesale we usually start with the draft quote from an emailed enquiry, because that is what eats the most hours in a salesperson's week. We take the next area only after the measurement on day 30, rather than straight away.

Will this make us sell more?

We do not promise higher sales or a better margin, because that is decided by the salesperson and the market, not by a workflow. We promise a shorter time from enquiry to quote sent and fewer mistakes in prices and item codes. If you win business on how fast you answer, this is exactly the stretch that can be shortened.

We also sell retail through an online shop and marketplaces. Is that a problem?

The opposite, it is a common setup in wholesale and usually a reason to start sooner. Retail and trade orders come out of the same warehouse, so stock has to be counted once rather than separately in every channel. We connect the shop and the marketplace feeds to the stock system so the salesperson sees one stock figure, not three.

We have no IT department. Who runs this afterwards?

We take the technical part, and after the rollout it is run by the person who knows the sales process, not by a developer. We teach them to change the things that change most often: price control thresholds, the wording of the confirmation, reminder timings. You only need to come back to us when the source system itself changes.

## Let us walk one of your enquiries end to end

The free consultation takes 30 minutes. We take one real enquiry out of your inbox and go through it step by step: who reads it, where the price comes from, how long the quote takes to build and what happens when the customer confirms. We say what can come off the salesperson right away and what needs the product records tidied up first. You do not have to give us access to any system at this stage.

We run at most two rollouts at a time, so that someone is really sitting on each one.

[Book a free consultation](https://justautomate.ai/en/contact/)

**PS.** If a salesperson re-keyed an order from an email this week, that is the cost we are talking about. We take the work after the conversation off him so that he can have more conversations. You write directly to Paweł, no reception and no tickets.

## Related

- [B2B sales automation](https://justautomate.ai/en/services/b2b-sales-automation/)
- [Automatic proposal sending](https://justautomate.ai/en/automations/automatic-offer-sending/)
- [Automatic inventory update](https://justautomate.ai/en/automations/automatic-inventory-update/)
- [Invoice automation](https://justautomate.ai/en/services/invoice-automation/)
- [BOND: business AI agent](https://justautomate.ai/en/services/bond-business-ai-agent/)
- [E-commerce automation](https://justautomate.ai/en/industries/e-commerce/)

---

**JustAutomate** · contact: contact@justautomate.ai · form: https://justautomate.ai/en/contact/