AI & the future of wholesale

AI demand forecasting works out how much of each item is needed, at every location, over the coming weeks. The model learns from order history and combines it with seasons, promotions, the weather and projects customers have announced. Much of the range sells only a few times a year, and one large order can throw any forecast off. AI picks up patterns a fixed rule of thumb misses and shows how certain each forecast is; the buyer decides on the exceptions. The next step looks beyond the wholesaler's own orders. Connected devices and supply cabinets at customer sites report what they need, and demand becomes visible before the customer is aware of it.

AI reads business orders from emails, PDFs and photos of handwritten lists, recognises items even when customers describe them in their own words, and enters the order with the right item number. It compares the order with the same customer's earlier ones and flags anything unusual, such as ten times the normal quantity or an item that has been discontinued. A member of staff only reviews the doubtful cases. Reading documents is a transitional phase, though. A growing share of orders will no longer arrive as a document at all, but directly from the customer's procurement agent or connected device, which keeps track of what is needed.

Linked to the wholesaler's own catalogue, product specifications, stock and customer history, a language model becomes an assistant for inside sales. Finding an alternative for a discontinued item becomes much faster. The interesting work is comparing and assembling. The assistant sets options side by side on price, quality, lead time and cost over their working life, then proposes the complete order: for a new restaurant kitchen, the equipment, tableware and cleaning products; for a building project, the materials and fixings. The member of staff and the customer choose what fits. With the assistant's help, a new colleague can put a proposal forward sooner.

An AI agent completes a task by itself within agreed limits: looking up data, making a decision and taking the next step. Wholesalers already use agents for routine daily work: answering a customer who wants to know when a delivery will arrive, preparing a standard quote for familiar items, chasing a quote still open after a week or an unpaid invoice. Fixed limits on amounts and discounts set what the agent may do alone; unusual questions and large customers go to a member of staff. The next step is an agent that handles a complete customer request, starting with the technical question and moving through quote and delivery to the invoice, dealing with the customer's own agent along the way.

A wholesaler buys from hundreds of suppliers, each with its own prices, lead times and minimum quantities. AI tracks how reliably each supplier delivers, spots lead times creeping up and estimates where the prices of raw materials such as copper and plastic are heading. Based on forecast demand, it suggests when to order, from whom and how much, weighing volume discounts against the cost of holding stock. When something runs short, it looks for alternative suppliers or substitute items. The next step: the wholesaler's procurement agents request quotes from suppliers themselves and negotiate within fixed limits. Buyers then concentrate on strategic suppliers and scarce product groups, where the difference lies in knowing who can still deliver in a tight market.

An AI-first wholesaler lets AI do the repetitive and predictive work that people need not do. Handling requests, drafting proposals and keeping data up to date go through AI wherever possible. Staff focus on the customer with a difficult problem, the question no manual answers and the negotiation with a large customer or supplier. AI runs the recurring calculations and produces forecasts. A scenario for 2040: the same team handles far more requests. The wholesaler becomes a partner to its customers, with staff discussing which products and delivery times fit their plans.

An account manager in wholesale often looks after dozens of customers, and not every customer gets the same attention. An agent follows all of them: orders, open quotes, complaints, payments and shifts in ordering behaviour. It prepares a visit with an overview and a few proposals, and settles small matters itself. The account manager holds the conversation in the customer's workshop and agrees the terms of a large project. Soon the account manager will also face a counterpart that is not human: the customer's procurement agent. The account manager's work then shifts to agreeing what that agent may buy from the wholesaler, and on what terms.

Leadership in wholesale means moving ahead while the daily orders keep coming in. Leaders take the initiative on AI and new technology and set out a picture of the future for the whole organisation. They encourage their people to help build it, and show experienced staff that their knowledge of customers and products is indispensable. Staff who have doubts also get room to have their say about the company's future. Leaders are honest about what is changing and dare to decide before everything is certain. That is how they build tomorrow's wholesale business together with their people.

Wholesale is a relationship business. Many staff have known their customers for years and know what they need before they call. That experience is valuable, and new ways of working ask staff to combine it with suggestions drawn from data. A sales rep who speaks to a customer every week does not immediately see why a system should suggest whom to call. Willingness often grows from what staff see change in their own work, such as a quote ready in five minutes rather than an hour. When testing, experienced people are often the first to see where the system misses how things work in practice, and their judgement is what makes it better.

A four-year course cannot keep up with the pace of AI. So organisations take learning into their own hands, in a future academy, without waiting for the education system. Current staff learn how AI and new technology change their work and how to work alongside agents. New staff practise with the organisation's own applications and ways of working. Leaders learn to make choices about work that is increasingly done by people and machines together. That way, the organisation builds its future with the people it already has.

In an ecosystem, an organisation draws much of its value from the parties around it. Partners bring practical knowledge, start-ups bring new technology and researchers study what it delivers. Governments help build shared facilities, such as digital trade documents and agreements on data. Alliances combine the strength of independent organisations and develop AI applications together. Across the border, organisations already have experience with agents and new business models; whoever learns from them moves ahead faster.