Insights from the Made in Group's November Marketers in Manufacturing Discussion Group.
Made in Group's most recent Industry Meetup saw the return of the new Marketing Sub-Group which tackled one of the fastest-moving challenges facing manufacturers today: how Large Language Models (LLMs) such as ChatGPT, Perplexity and Gemini are reshaping the way buyers find and evaluate suppliers.
Unlike traditional search engines, LLMs don’t list websites — they summarise a business. And if the model can’t understand your company clearly, it simply won’t recommend you.
This shift framed a lively and highly practical discussion among members from across Made in Group's manufacturing community.
Who Was in the Room?
The discussion group comprised:
Chair:
-
Sam Sleight (Made in Yorkshire)
Delegates:
- Dan Ireland (Garthwest)
-
Jacqueline Watkins (Hyfore Workholding)
-
Emma Hockley (Big Bear Plastics)
-
James Miller (LK Metrology)
-
Georgette Donoghue (Ellis Patents)
-
Michelle Mason (Stage One Creative Services)
-
Chris McGuiness (Toolex)
- Jackie Cook (Airmaster)
-
Phil Crothers (RamBase)
-
Bret Towler (AML Instruments)
-
Vrinda Gupta (Powdertech Surface Science)
-
Chloe Talbot (Trust Electric Heating)
-
Leonie Husselbee (Trust Electric Heating)
-
Simon Price (Power Plastics)
- Karen Tinkler (Partners PR)
-
Mariam Taha (Fluere)
-
Jonathan Wright (Colchester Machine Tool Solutions)
Risk, Opportunity — or Both?
To get a top-level overview of the perception of AI, the session opened with a rapid-fire question from Sam Sleight of Made in Yorkshire: is AI primarily a risk or an opportunity for your organisation?
Not one participant chose “risk only”.
Most saw significant opportunity — particularly in productivity, discoverability and efficiency — but several, including Partners PR and Airmaster, noted that the benefits come with challenges around accuracy, confidentiality and data security. Big Bear Plastics reflected that opportunity rarely exists without risk. RamBase emphasised the potential dangers of employees unintentionally exposing sensitive information.
What became clear is that AI is already influencing buying behaviour. When Sam asked who had used AI to support a personal or business buying decision, roughly half the room raised their hands — a trend everyone agreed would only accelerate as younger procurement professionals come through.
What AI Actually Reads — and Why It Matters
Sam Sleight shared key insights from his research: LLMs don’t “rank” websites like Google. Instead, they build a picture of a company using information pulled from:
-
Websites
-
LinkedIn
-
Directories
-
Job ads
-
Reviews
-
News articles
-
Case studies
From there, LLMs assess two things:
How clear and consistent the content is, and how credible the business appears across multiple sources.
This resonated strongly with the group, particularly when James Miller of LK Metrology shared a real-world example.
A few months ago, LK Metrology weren’t appearing in ChatGPT’s suggested suppliers at all, while a much younger competitor was. That discovery prompted a full audit of their outward-facing content, tighter alignment across channels and a renewed focus on clear, recent case studies — especially UK examples featuring customer quotations.
When they re-tested, their visibility improved.
James also shared his 20-60-20 approach to working with AI:
20% clear inputs → 60% AI drafting → 20% human refinement.
It’s a balanced approach that captures the efficiency gains LLMs offer for content creation, while still preserving the human insight and nuance that remain so important — and increasingly valuable.
Proof Points: Credibility & Clarity
Credibility
A strong theme in the discussion was the importance of publicly visible, real-world proof…
Members agreed that LLMs favour authentic, evidence-rich content over polished marketing language. Certifications, testimonials, technical detail and up-to-date case studies were recognised as essential trust signals.
Additional testing carried out by Made in Yorkshire reinforced this point. When analysing MIY’s digital footprint, ChatGPT highlighted a testimonial-style reference on Dale Power Solutions’ website as part of its reasoning — showing how positive external mentions can strengthen perceived credibility. Alignment with recognised industry networks also contributes to this, as LLMs draw on contextual associations when forming trust judgements.
Practical realities also surfaced. As Jackie Watkins (Hyfore) explained, NDAs often limit the ability to publish imagery or detailed project descriptions. This makes structured written proof especially valuable. Hyfore saw a tangible benefit after rebuilding their CNC capability pages into consistent tables with clear specifications and alt-texted images — a format that finally helped them appear for subcontract machining work they have long delivered.
Collectively, the group recognised that LLMs reward evidence, consistency and coherence across platforms. And that naturally led into the second theme of the discussion: not just what you publish, but how clearly you express it.
Clarity
The second thread of the conversation centred on how AI interprets language. Members agreed that overly marketing-led or vague messaging can actively confuse LLMs.
What AI models — and procurement teams — value is straightforward:
-
Plain-English explanations of what you make
-
Clear descriptions of capabilities
-
Defined tolerances, materials and certifications
-
Clear sector alignment
- Detailed FAQ Page
-
Facts, numbers and evidence rather than adjectives
This theme linked closely to Hyfore’s example. Their previous website had been visually engaging but too unstructured for AI to understand. By reorganising content into consistent, factual formats, they made life easier for both buyers and LLMs.
The message that emerged from the group was simple but powerful:
Clarity beats cleverness.
Manufacturers that articulate what they do in unambiguous, structured terms give AI systems the best possible chance of understanding — and accurately representing — their capabilities.
Tools Members Are Using
Several members shared practical examples of where AI is already supporting their marketing and operational work.
Sintra – Trust Electric Heating
Chloe Talbot and Leonie Husselbee explained how Trust Electric Heating are using Sintra, a platform built around multiple specialist AI bots that support SEO, content and website optimisation. While rebuilding their website, they’ve found Sintra particularly useful for:
-
SEO support via Siomi, which helps them understand how well their content aligns with search intent.
-
Content creation, using bots that generate blog ideas, web copy and social media posts with relevant trends and hashtags.
-
Website diagnostics, with Vizzy scanning their existing site and providing a clear, step-by-step breakdown of what needs improvement before the new version goes live.
For Trust, Sintra acts as both a creative assistant and a technical auditor — helping them move towards a more discoverable, AI-friendly online presence.
Microsoft Copilot – Power Plastics
Simon Price offered a different but equally practical example. Power Plastics are exploring how Microsoft Copilot, connected securely to a SharePoint library of tenders, certificates and past RFP responses, can support the efficiency of their bid and compliance processes. Early testing showed Copilot could quickly pull together accurate draft answers to questions such as ISO credentials — provided their documentation is up to date and human review remains central.
While not directly linked to SEO, this example highlighted a wider point echoed by the group: AI’s value in manufacturing goes far beyond marketing, with clear opportunities to streamline internal workflows when used responsibly.
In Summary
The discussion made one thing abundantly clear:
AI isn’t replacing search — it’s redefining discoverability.
Manufacturers who communicate clearly, present structured content and maintain consistency across platforms are the ones AI systems feel confident recommending.
As Sam Sleight of Made in Yorkshire noted, buyers are increasingly asking AI tools: “Who can help me with this?”
Whether the model recommends you depends entirely on the story your business tells online — and how clearly it can be understood.