How AI can give workplace salespeople the confidence to diversify

Gary Naphtali – This article first appeared in the September 2026 edition of Workplace360 magazine 

THE PROBLEM ISN’T THE PRODUCT

I have spent most of my working life in and around the workplace supplies sector, and I have watched it reinvent itself more than once. We moved from product catalogues and separate suppliers to single-source solutions. We adapted to e-commerce, consolidation, Amazonification, hybrid working and a fundamental shift in what the office actually is. Now we are being asked to reinvent ourselves again – this time through diversification.

The logic is difficult to argue with. Traditional office supplies are a mature, declining category. Customers still need them, of course, but they do not occupy the same share of budget, attention or strategic importance that they once did. Meanwhile, the wider workplace presents a substantial opportunity: furniture, interiors, hygiene, washroom, catering, workwear, print, technology, managed services, sustainability and many other categories that sit naturally around the customer we already serve.

Yet the pace of adoption remains mixed. Some dealers and salespeople have embraced the broader workplace proposition with energy. Others have added categories to a catalogue or website but have not truly added them to the sales conversation. The product is technically available, but the salesperson still opens with paper, toner and stationery.

I do not believe this is usually laziness or a lack of ambition. In many cases, it is confidence. An experienced salesperson may know office supplies inside out. They understand the products, the margins, the likely objections and the people who buy them. Ask that same person to lead a conversation about washroom compliance, workplace design, sustainability or managed print and suddenly they feel like a novice. Retreating to what they know is entirely human.

The danger is that experience becomes an anchor. The salesperson waits until they know everything about a new category before mentioning it; the customer continues buying it elsewhere; and the dealer concludes there is little demand. That is not a product failure. It is a confidence and positioning failure.

THE OPPORTUNITY HIDING IN PLAIN SIGHT

The irony is that workplace salespeople are better placed than almost anyone to diversify. You already have relationships, trading history, service credibility and access to the customer’s workplace. You may know where every ream of paper is delivered, how many sites the organisation operates, when its busiest periods occur and which people influence purchasing. A new entrant would pay dearly for that intelligence.

Diversification does not require you to become the world’s leading expert in every category. It requires you to become expert at recognising where a customer could benefit, asking better questions and bringing in the right specialist when required. Your value is not limited to knowing the product code. It lies in understanding the customer’s operation and joining the dots.

This is where AI can make a genuine difference. Not by replacing the salesperson, writing thousands of generic emails or pretending to know things that it does not. Its most useful first role is much simpler: helping the salesperson prepare. It can reduce the blank-page fear, organise scattered information, reveal useful patterns and suggest questions that move a conversation beyond office supplies.

For an established salesperson, that creates a bridge between the confidence of the old category and the opportunity of the new one.

‘I KNOW, NOT I THINK’

One of the principles I use in sales training is ‘I Know, Not I Think’. It means we should not approach a customer merely because we think they might buy something. We should engage when we know – based on evidence – that we can improve some aspect of their business.

That does not mean knowing the final answer before the first conversation. It means doing enough work to form a credible, testable point of view. Instead of saying, ‘I wondered whether you might be interested in our new furniture range’, you should be able to say, ‘I believe we can improve your workplace in three specific areas, and here is the evidence that led me to that conclusion.’ Better still, earn the right to say: ‘I know we can help – let me show you why.’

Those reasons might be that the organisation is recruiting quickly and will need to equip new people; it has opened another site and appears to be managing suppliers locally; its sustainability commitments are not reflected in its current purchasing approach; or its offices were designed for a working pattern that no longer exists. None of those insights starts with a product. They start with the customer.

In the past, gathering that level of intelligence could take hours. Today, AI-supported research can do much of the heavy lifting in minutes – provided the salesperson directs it properly and checks the result. The tool does not create the commercial judgement. It gives your judgement more material to work with.

START WITH THE CUSTOMER, NOT THE TOOL

The worst way to begin is to buy an impressive AI platform and then ask the sales team what to do with it. Start with a customer, a commercial question and a defined outcome. Choose one existing account where you believe there is untapped potential, or one prospect that closely resembles your best customers. Then use the following process.

1. BUILD A USEFUL ICP

Most ideal customer profiles are too shallow. ‘Companies with 50–500 employees within 50 miles’ is a list criterion, not an ICP. A useful ICP describes the organisations most likely to value your particular solution.

Begin with three to five customers where you already deliver strong results and acceptable margin. Give an AI assistant non-confidential facts about them: sector, size, number of sites, workplace model, growth stage, buying complexity, service challenges, categories purchased and the outcomes you create. Ask it to identify the common characteristics, likely trigger events and operational problems that make those customers a good fit.

The result should not be accepted as gospel. Discuss it with colleagues and compare it with what you know. The exercise is valuable because it converts instinct into a working hypothesis. You can then ask: which other organisations share these characteristics, and what evidence would confirm the fit?

2. CREATE A CUSTOMER INTELLIGENCE BRIEF

Select one target account and gather information from reliable public sources: its website, annual report, Companies House filings, job adverts, LinkedIn activity, local and trade press, sustainability statements, office or site announcements, leadership changes and published supplier commitments. Search-enabled AI tools such as ChatGPT or Microsoft Copilot can help bring this material together and summarise it. LinkedIn Sales Navigator can add account, people and trigger-event insight. Companies House remains a useful free source for filings, officers, addresses and company history.

Ask the AI to organise what it finds under commercially useful headings: what is changing; likely workplace implications; possible cost, risk, efficiency or employee-experience pressures; relevant decision-makers; evidence still missing; and questions that must be verified directly with the customer.

Insist that it shows its sources. If a claim cannot be traced to a credible source, label it as an assumption or remove it. AI can be confidently wrong, and a salesperson repeating an invented fact will destroy the very credibility this exercise is intended to build.

3. LOOK FOR CHANGE – BECAUSE CHANGE CREATES NEED

A static company description rarely produces a compelling reason to call. Trigger events do. Recruitment, acquisitions, new sites, relocations, contract renewals, new leadership, sustainability targets, hybrid-working changes, regulatory pressure and rapid growth all create new workplace requirements.

Ask the AI: ‘What has changed in this organisation during the past 12 months, what workplace consequences could follow, and which of our solution areas might be relevant?’ Then challenge every answer. Is the link logical? Is it supported? Is it important enough for the customer to act upon?

Imagine a regional legal practice recruiting 40 people across three offices. The obvious office-supplies response is to quote for more stationery. The solution-led response is to consider furniture and ergonomics, onboarding packs, print and document workflows, hygiene, refreshments, storage, branded workwear where appropriate, and the consistency of supply across sites. AI has not sold anything. It has helped the salesperson see the whole workplace.

4. TURN FEATURES INTO BUSINESS BENEFIT

New-category confidence often collapses because the salesperson learns product features rather than customer outcomes. Features matter, but they rarely start a senior-level conversation.

Give the AI a short description of your solution and ask it to translate each feature through three levels: the immediate benefit, the operational effect and the wider business outcome. A consolidated facilities-supplies programme might reduce the number of orders. That could reduce invoice processing and management time. The time released might allow a finance or operations team to focus on more valuable work. That final ‘sell-through benefit’ is usually where the strongest conversation begins.

Ask it to produce benefits for different stakeholders too. A finance director, facilities manager, HR leader, sustainability lead and end user will not value the same thing in the same way. You are not changing the truth; you are making the relevance visible.

5. PREPARE THE THREE REASONS

Before making contact, complete one sentence: ‘I know we can improve this customer’s business because…’ Then add at least three evidence-backed reasons.

For example: ‘You have opened two additional sites, but your published information suggests local purchasing is still decentralised. You have committed to reducing supply-chain emissions, yet your current model may involve multiple deliveries and vendors. You are recruiting heavily in operations, which suggests that standardised workplace onboarding and replenishment could release management time.’

Notice the language: specific, commercial and centred on the customer. Also notice the restraint. ‘Suggests’ and ‘may’ distinguish an informed hypothesis from a known fact. The purpose of the first approach is not to announce that you have diagnosed the business from the internet. It is to show that you have done enough work to ask an intelligent question.

6. USE AI TO REHEARSE – NOT TO HIDE

Once the case is clear, ask the AI to act as the customer. Tell it the person’s likely role and ask it to challenge your three reasons. What would a sceptical finance director question? What evidence would a facilities manager need? Which part sounds generic? What might make the proposal feel risky?

Then practise answering aloud. Ask the tool to score whether your opening is customer-centred, evidence-led and free of unnecessary jargon. This is particularly helpful when entering an unfamiliar category: you can test the conversation privately before testing it with the client.

Finally, use AI to draft a short approach, but never surrender your voice to it. If the result sounds as though it could have been sent to 500 companies, it probably will not interest one. Add the detail only you know: the relationship history, the operational frustration mentioned in a meeting, the site you visited, or the colleague who should be involved. AI should increase personalisation, not industrialise blandness.

A SIMPLE FIRST-WEEK CHALLENGE

You do not need a transformation programme to begin. Choose one salesperson, one account and one adjacent category. Allow 45 minutes for structured research. Produce a one-page brief containing five items: what has changed; what the customer may need; the evidence; three questions to verify; and three reasons you can create benefit. Review it with a manager or category specialist, make the approach and record what happens.

Repeat the exercise with five accounts. You will quickly learn which sources are useful, which prompts produce noise and where specialist knowledge is required. Save the best prompt and brief as a team template. The habit matters more than the technology.

A useful starter prompt might be: ‘Act as a commercial research assistant for a UK workplace-solutions salesperson. Using current, cited public information, analyse [company]. Identify recent changes that could affect its workplaces, employees, sites, supply chain or purchasing. Suggest possible implications across [our relevant categories]. Separate verified facts from hypotheses. For each hypothesis, provide the evidence, the likely stakeholder and one question I should ask to validate it. Do not invent information.’

That final instruction matters. So does your company’s policy on data. Do not upload confidential customer information, pricing, contracts, personal data or commercially sensitive material into an unapproved public tool. Use company-approved systems and keep a human responsible for every conclusion and every communication.

THE TOOLKIT CAN BE SMALL

For the first steps, I would resist assembling a long list of expensive platforms. One search-enabled AI assistant, your CRM, LinkedIn, Companies House, the customer’s own website and a disciplined research template are enough to create a meaningful improvement.

If your organisation already uses Microsoft 365, explore the approved Copilot options available to you. If you use ChatGPT, use web search or deeper research features when current, sourced information matters. If prospecting is a major priority, Sales Navigator can help identify people, relationships and changes within target accounts. Whichever tool you choose, its value depends on the question you ask and the judgement you apply afterwards.

The salesperson remains accountable for checking facts, protecting data, understanding nuance and deciding whether the insight is commercially relevant. AI is a capable research assistant; it is not a witness, an expert in your customer or an excuse to stop thinking.

CONFIDENCE FOLLOWS PREPARATION

The workplace supplies sector does not lack products. It does not lack supplier partnerships, catalogues or categories. What it sometimes lacks is the confidence to begin a different customer conversation.

Experienced salespeople possess something AI cannot manufacture: years of pattern recognition, commercial instinct, resilience and trusted relationships. AI can help them apply that experience more widely. It can accelerate research, expose connections and help turn an unfamiliar category into a credible business discussion.

Do not begin by asking, ‘How can I sell this new product?’ Ask, ‘What is happening in my customer’s world, and where can we make it better?’ Then gather the evidence until your proposition is no longer a hopeful claim.

Diversification will not be won by the dealer with the longest catalogue. It will be won by the salesperson who understands the customer best, sees the opportunity first and is prepared to make a position statement based on knowledge.

Move from ‘I think they might need this’ to ‘I know we can help – and here are three reasons why.’ That is a far better place to start.