This site uses cookies for analytics and to improve your experience. By clicking Accept, you consent to our use of cookies. Learn more in our privacy policy.

Lets Talk

Category: Promotional Merchandise

  • From Chatting to Doing: The One Thing to Get Right Before You Let AI Take Action in Your Business

    From Chatting to Doing: The One Thing to Get Right Before You Let AI Take Action in Your Business

    Most of us have now had the first AI experience: you ask ChatGPT (or Claude, or Copilot) to draft an email, summarise a supplier price list, or explain what “DTF printing” actually is, and it does a decent job in seconds. That’s AI as a chat assistant, acting as a very well-read colleague you can bounce things off.

    The next step is much more interesting, and it’s already happening: AI as an agent, software that doesn’t just tell you things, but actually does things. Drafts the quote and drops it into your system. Chases the artwork approval. Checks stock with the supplier and updates the order. For a promotional merchandise business drowning in repetitive admin, that’s where the real time savings live.

    But there’s a catch, and it’s worth understanding before you make the jump. It comes down to one unglamorous word: grounding.

    Why the jump from chatting to doing raises the stakes

    When AI is just chatting, a wrong answer costs you a few minutes. You read it, think “that’s not right,” and move on. When AI is acting, a wrong answer becomes a wrong quote, a wrong order, or a promise your business is now expected to keep.

    That’s not a hypothetical. In 2024, a Canadian tribunal ordered Air Canada to compensate a passenger after its website chatbot invented a bereavement fare policy that didn’t exist — and rejected the airline’s argument that the chatbot was “a separate legal entity responsible for its own actions.” If your AI tells a customer something, you own it.

    It gets more expensive as the work gets more important. In 2025, Deloitte partially refunded the Australian government after a AU$440,000 report was found to contain fabricated academic references and a made-up quote from a court judgment. AI-generated content that nobody had properly checked. And in one widely reported case the same year, an AI coding agent at Replit went a step further than writing anything at all: it deleted a company’s live database during a code freeze, prompting a public apology from the company’s CEO. In the legal world, a tracker maintained by researcher Damien Charlotin has now logged more than 1,800 court cases worldwide in which lawyers were caught filing documents containing AI-fabricated citations.

    Different industries, same underlying story: AI output that sounded confident, wasn’t checked against anything real, and was allowed to reach the point where it had consequences.

    Why AI makes things up in the first place

    This isn’t AI being careless, and it isn’t a bug that next year’s model will simply fix. Large language models work by predicting the most plausible next words based on patterns in their training data. Researchers at OpenAI published a paper in 2025 explaining that models hallucinate partly because of how they’re trained and tested: the process rewards confident guessing over admitting uncertainty  just like a student who never leaves an exam answer blank, because a wrong guess scores better than “I don’t know.”

    So a raw language model asked “what’s our trade price on 500 printed mugs?” will happily produce a plausible-sounding number. Plausible is what it’s built for. Correct requires something extra.

    Grounding: giving the AI an open-book exam

    That something extra is grounding. Grounding means connecting the AI to real, verifiable sources of information such as your price lists, your product data, your terms of business, a supplier’s live stock feed and requiring it to base its answers on what it finds there, ideally with citations showing exactly where each claim came from. Google and Microsoft both now build grounding and “groundedness checking” into their enterprise AI platforms, which tells you how central it has become.

    The best-known grounding technique is retrieval-augmented generation (RAG), first described by researchers in 2020. The idea is simple: instead of asking the model to answer from memory, the system first retrieves the relevant documents, the right page of the right price list and then hands them to the model along with your question. It turns a closed-book exam into an open-book one. The model stops guessing what your prices probably are and starts reading what they actually are.

    For a business, grounding buys you three things. Answers based on your data rather than the internet’s averages. Citations, so a human can check where a claim came from. And a system you update by updating your documents; so new price list in, new answers out – with no mysterious retraining involved.

    There’s a caveat: grounding reduces errors, it doesn’t abolish them

    Here’s the part vendors tend to whisper. A Stanford University study tested premium legal AI research tools built on RAG, products whose marketing claimed to “avoid” hallucinations and found they still produced incorrect or misgrounded answers between 17% and 33% of the time. That was a big improvement on general-purpose chatbots, but a long way from perfect. If your knowledge base contains an out-of-date price list, a grounded AI will faithfully cite the wrong price with a lovely footnote.

    The practical conclusion isn’t “don’t bother.” It’s that grounding is necessary but not sufficient but it needs to be paired with sensible checkpoints. Which brings us back to the bridge between chatting and doing.

    Crossing the bridge safely

    The move from chat assistant to acting agent doesn’t have to be a leap. Sensible businesses cross it in stages:

    1. Grounded answers first. Start with an assistant that can only answer from your own documents, and shows its sources. This is where you find out how good (or stale) your product data, pricing and policies really are.
    2. Drafts, not deeds. Next, let the AI prepare actions — draft the quote, pre-fill the order, write the chase-up email with a human clicking “send.” You get most of the time savings while a person still owns every consequence.
    3. Narrow autonomy, with limits and logs. Only once a task has proven itself do you let the AI complete it end-to-end — and even then, within tight boundaries (order values it can handle, actions it can never take) and with a log of everything it did and which sources it relied on.

    Notice what makes stages two and three possible at all: the grounding you built in stage one. An agent is only as trustworthy as the information it acts on. Get the plumbing right with your data, connected, current and citable. Then the exciting stuff stops being a gamble.

    AI that chats is a better search box. AI that acts is a new member of staff. And like any new member of staff, it should be given the company handbook, access to the real files, clear limits on what it can sign off, and a manager who checks its work until it’s earned trust. That’s grounding, and it’s the difference between AI that impresses in a demo and AI you’d actually let near your customers.

    Sources: CBC News — Moffatt v. Air Canada · The Register — Deloitte Australia refund · Fortune — Replit database incident · Damien Charlotin — AI Hallucination Cases tracker · OpenAI — Why Language Models Hallucinate (2025) · Lewis et al. — Retrieval-Augmented Generation (2020) · Google Cloud — Grounding · Microsoft — Groundedness detection · Stanford RegLab — Hallucination-Free? (2024)

  • Sourcing Unique, Valued Items on Limited Time, Skills and Budget

    Sourcing Unique, Valued Items on Limited Time, Skills and Budget

    Finding distinctive, well-received merchandise does not require a dedicated sourcing team or a big budget, but it does require a smarter process than scrolling a catalogue.

    Most people responsible for buying promotional merchandise are not merchandise specialists. They are marketers, office managers, HR leads or founders who have been handed a task alongside a dozen others, a limited budget, and an expectation that whatever they choose will not feel like everyone else’s giveaway.

    The instinct in this position is to search a generic catalogue site, filter by price, and pick whatever looks reasonable within the first page of results. The problem is that everyone else under the same time pressure does exactly the same thing, which is why so much promotional merchandise looks and feels identical: the same mugs, the same tote bags, the same USB sticks that quietly went out of fashion years ago.

    Finding something distinctive on limited time does not require becoming a sourcing expert. It requires a slightly different process. Start with the recipient, not the product. Who is receiving this, and what do they already value, use or aspire to? A small, specific insight (a team that cycles to work, a client industry with a particular seasonal rhythm, an audience that cares visibly about sustainability) narrows a search far faster than browsing by category.

    Next, resist the temptation to search only for “promotional” versions of a product. Some of the best-received branded items began life as a well-made, ordinary product from a smaller maker that happens to also offer branding, rather than a product designed from the outset to be given away. These items tend to read as considered rather than generic, because they were not built to be cheap in volume.

    Time and skill constraints are real, though, and this is precisely where artificial intelligence tools are starting to earn their keep in this industry. Rather than manually trawling supplier catalogues, AI-assisted search can match a brief (budget, audience, values, quantity, timeframe) against a much wider supplier base in minutes rather than hours, surfacing options a busy buyer would never have found by scrolling.

    A modest budget and no formal sourcing background are not, on their own, a barrier to finding something genuinely well received. What actually limits most buyers is time, and that is the constraint worth solving for first, whether through a sharper brief, a smaller and better-chosen supplier list, or tools that do the trawling for you.

    Six Different Ways builds AI-powered sourcing tools that do exactly this kind of matching for distributors and buyers. If catalogue trawling is eating time you do not have, it is worth a conversation.

    Woman browsing promotional materials on desktop computer
    A professional searches online for promotional materials at her desk. The bright screen contrasts with the calm, sketch-style setting.

  • What If Your Clients Had a Merchandise Expert Available at 11pm?

    What If Your Clients Had a Merchandise Expert Available at 11pm?

    AI advisors are quietly becoming a competitive advantage in promotional merchandise — handling the routine questions so your team can focus on the ones that matter.

    Ordering promotional merchandise sits somewhere between an enjoyable creative task and a mildly stressful procurement one. The enjoyable bit is choosing what to make. The stressful bit is everything else — minimum order quantities, lead times, decoration methods, artwork specs, sustainability credentials, budget trade offs.

    The people best placed to answer those questions are the suppliers and distributors clients already work with. The problem is those conversations happen on the supplier’s schedule, not the clients. A question that comes up at 9pm on a Tuesday, or in a planning meeting where no external party is present, just sits there until someone finds time to send an email and wait.

    This is exactly the gap AI conversational tools are starting to close — and the best implementations in this industry are doing something genuinely useful, not just gimmicky.

    A well-configured AI advisor, deployed by a supplier or distributor, can field a surprising range of the questions that currently sit unanswered in inboxes. What’s the minimum order for embroidered polo shirts? How long will production and delivery take? What’s the difference between screen print and digital print for this particular design? Is this product available in recycled materials? None of these require the creative or strategic judgement of a skilled account manager — they just require accurate, fast answers.

    From the client’s side, this is something they’re already used to in plenty of other areas of their working life. ChatGPT has set the bar for instant, knowledgeable help, and clients now expect something similar from the businesses they buy from. A supplier offering this experience stands out — not because the tech is novel anymore, but because it removes friction at exactly the moments that matter.

    It’s worth being honest about the limits though. AI advisors aren’t suited to complex strategic conversations about a brand’s overall merchandise approach. They can’t read the room or pick up on the unstated subtext an experienced account manager handles instinctively. For big, complicated briefs, the human conversation is still essential — and the best AI advisor implementations are designed to hand off gracefully rather than bluff an answer they’re not equipped to give.

    Get this balance right and you’re not replacing your best people. You’re freeing them up to spend their time on the work that actually needs them.

  • The Best Time to Talk to a Client Is Before They Start Looking

    The Best Time to Talk to a Client Is Before They Start Looking

    AI can tell you exactly when that moment is for every client on your books — not just the ones you happen to remember.

    Most promotional merchandise businesses operate reactively. A client calls, explains what they need, and the process kicks off. The trouble with this model is that by the time a client picks up the phone, there’s a decent chance they’re already talking to two or three other suppliers. You’re starting from zero, and starting late.

    The businesses that consistently win more than their fair share tend to be the ones who show up earlier — sometimes before the client has even fully worked out what they need. Historically that’s been down to experienced account managers with a good feel for the calendar and the client. Valuable, sure, but it doesn’t scale and it walks out the door the day that person leaves.

    AI offers a more systematic route to the same outcome, and it’s more accessible then most people assume.

    Start with what you already have: order history. Every client has a pattern, even if nobody’s written it down. Annual conference merchandise ordered every September. Christmas gifts briefed by late October. AI tools can scan years of purchase history, spot these recurring windows, and flag which clients are approaching theirs — giving your account managers the prompt to reach out before a competitor does.

    Beyond your own data, there are external signals worth watching too. A client posts about a major product launch on LinkedIn — that’s a strong hint they’ll need event merchandise soon. They post five new sales job listings — onboarding kits and welcome packs are probably coming. They rebrand — almost every piece of branded merchandise they own just became outdated overnight, and that’s your moment. AI monitoring tools can track these signals across your whole client list rather than relying on someone happening to notice.

    There’s a sector-level version of this too. If you work with retail clients, you already know their promotional calendar. If you work in financial services, you know roughly when their corporate gifting season runs. Building a proactive outreach calendar around these known rhythms means the right clients hear from you at the right moment, rather than receiving a generic newsletter whenever someone gets round to writing one.

    None of this is about replacing your best account managers — it’s about giving every client the same level of proactive attention, regardless of how stretched the team is or how new the relationship happens to be.

  • Your Sales Reports Are Full of Answers. You’re Just Not Asking The Right Questions

    Your Sales Reports Are Full of Answers. You’re Just Not Asking The Right Questions

    Most promotional merchandise distributors have more sales data than they know what to do with — and AI is the cheapest way to finally use it.

    Every distributor we speak to has a sales report. Most have several — CRM exports, order histories, margin breakdowns by client. The honest truth is that most of this data gets glanced at once a month and then forgotten about until the next one’s due.

    That’s not a criticism, it’s just how things tend to go when manually digging through spreadsheets takes hours nobody really has. The report ends up being a record of what happened rather then a tool that shapes what happens next.

    Here’s where AI genuinely changes the maths. A conventional report tells you Client A spent £12,000 last year. An AI-assisted analysis tells you Client A’s spend has dropped 23% over three quarters, that the decline is concentrated in gifting, their last order was eight months ago, and based on past patterns they’re overdue a phone call. That’s not a small difference — it’s the difference between data and direction.

    None of this requires a data science team or expensive software. Microsoft Copilot inside Excel can already analyse a spreadsheet and answer plain-language questions about it. ChatGPT’s Advanced Data Analysis does the same with an uploaded CSV. If you’re already using HubSpot or Salesforce, there’s likely an AI layer sitting there waiting to flag at-risk accounts — most people just haven’t switched it on yet.

    The barrier really isn’t the technology. It’s knowing what to ask, and building the habit of asking it regularly rather than once a quarter when someone remembers.

    A good starting list of questions: which clients haven’t ordered in ninety days, and what did they used to spend? Which product categories are growing as a share of total revenue? Which clients reliably order in Q4, and have they been contacted yet this year? Run these monthly and you’ll start to see patterns that a glance at a spreadsheet simply won’t surface.

    Distributors who do this well share a few habits. They’ve picked a small number of high-value questions rather than trying to boil the ocean. They keep their data structured consistently so AI tools can actually work with it. And critically, they translate every output into a specific action — a phone call, a reactivation email, a product recommendation — rather than leaving interesting observations to gather dust.

    The pay off tends to show up in two places: clients you’d otherwise have quietly lost get re-engaged, and upsell opportunities that were hiding in plain sight finally get acted on.

  • Why Your AI Mock-Up Might Not Be Printable (And What To Do Instead)

    Why Your AI Mock-Up Might Not Be Printable (And What To Do Instead)

    AI-generated mock-ups look stunning. The trouble is, a fair few of them describe products that can’t actually be made.

    If you’ve used Midjourney, ChatGPT or any of the popular AI image tools to picture what your branded mug, tote bag or water bottle might look like, you’ve probably been impressed by how real it looks. Crisp logo, perfect placement, flawless colour. There’s just one problem: a lot of these images show products that aren’t physically possible to produce.

    This is becoming a genuinely common issue across the promotional merchandise industry. Brand managers arrive at briefing meetings with an AI mock-up already in hand, fully expecting it to be exactly what gets delivered. The mock up usually ignores things like handle placement on mugs, print margins on garments, or the minimum stroke width needed for embroidery to actually hold its shape. AI doesn’t know — or care — about manufacturing constraints. It just knows what looks convincing.

    So why does this keep happening? Partly because the tools are improving fast, producing genuinely photorealistic results. And partly because procurement and marketing teams are under pressure to move quickly, and an AI mock-up is a fast way to get a concept on the table before anyone’s checked whether it’s producible.

    The good news: AI is also the best tool for solving this problem, just used a different way.

    Suppliers and distributors who get ahead of this are using AI to pre-qualify briefs before they reach the artwork stage. A constraint-aware chatbot, trained on your actual production specifications, can walk a client through what’s achievable before expectations are set in stone. Other businesses are using computer vision tools to flag obvious issues automatically — a logo too close to a handle, text below minimum readable size — in seconds rather than after a multi-day revision cycle.

    There’s also a more creative fix. Rather than simply rejecting a client’s AI concept, you can use AI to regenerate a compliant version — one that respects the real print area but keeps the spirit of what the client liked. It’s a far better experience than a flat no, and it tends to shorten the sales cycle considerably.

    If you’re a brand buyer reading this rather than a supplier: the practical takeaway is to treat any AI-generated mock-up as a starting point for a conversation, not a finished spec. Ask your supplier early whether what you’ve pictured is actually printable on that product, in that position, at that price. It’ll save a revision cycle later, and it tends to get you to a better result faster anyway.

    For suppliers, the arrival of AI mock-ups isn’t a problem to manage away — its a signal that the briefing process itself needs to evolve to keep pace with what clients now expect to see before they commit.