“Show Me, Don’t Tell Me”: AI in Manufacturing Needs Action, Not More Empty Promises

2025 can’t be another year of buzzwords and vague promises. To accelerate AI adoption, manufacturers need results-driven solutions that solve the problems they face every day.

It’s time to move beyond marketing slogans and demonstrate how AI can deliver real, measurable value, says XpertRule’s Akeel Attar and Iain Crosley.

We’ve all heard it, “AI will transform manufacturing.” AI is the future.” AI is going to change everything.” Well, guess what? Enough with the bravado. AI is not a magic cure for all your problems. It won’t instantly make your factory smarter, faster and more efficient. It won’t automatically engage workers or boost productivity overnight.

The problem isn’t AI’s potential, it’s the unsubstantiated hype that leaves decision-makers with solutions that underperform in real-world settings. Yes, AI can be groundbreaking, but it’s not a one-size-fits-all solution. It’s a tool. One that can be incredibly powerful when applied correctly but one that can also backfire if used in the wrong way.

Too many so-called “AI-powered” systems are little more than rebranded automation tools, with no explanation of how AI was involved in product development or how its inclusion drives added value. Without clear, demonstrable results, such claims risk eroding confidence and slowing the pace of meaningful adoption.

At XpertRule, we stand by the principle of ‘Show Me, Don’t Tell Me.’ Manufacturers deserve more than abstract promises, they need clarity and a clear plan for what problems AI can actually solve. But cutting through the noise requires asking the right questions – and that’s difficult without the right knowledge.

Knowledge Gap is Holding AI Deployment Back

More manufacturing companies than ever are exploring AI, with applications spanning predictive maintenance, inventory management, quality control, energy optimisation and customer service.

Yet, despite widespread enthusiasm, only 7% of manufacturers consider themselves highly knowledgeable about AI applications in their industry. That might explain why one in three say they use AI but fewer than one in ten (7%) have implemented it successfully. With most manufacturers (74%) planning to increase AI investments this year, addressing this knowledge gap must be a priority.

With a firmer grasp of what AI can (and can’t) do, decision-makers can ask the right questions, challenge vague claims and assess real value. Without it, they risk investing in systems that merely add noise to the conversation.

Key Questions to Ask AI Vendors

To separate genuine AI solutions from overhyped pitches, manufacturers should insist on clear evidence. Here are five direct, no-nonsense questions to ask:

1. Can you show me real-world results?
   Request specific case studies demonstrating how your AI system has improved outcomes in a production environment.
2. How does your AI improve on existing systems?
   Clarify how exactly AI improves decision-making, efficiency or accuracy beyond what standard automation tools can offer.
3. What data does your AI model need – and what happens if I don’t have it?
   Since AI’s effectiveness depends on high-quality, structured data, determine whether your existing infrastructure meets these needs.
4. How long until I see measurable improvements?
   Most manufacturers can’t afford long-term science experiments, so ask for a realistic timeline for delivering a return on investment.
5. What support is in place if issues arise?
   Ask what ongoing support, training and safeguards are provided to maintain performance and mitigate risks.

A Clear Path Forward

At XpertRule, we’ve seen firsthand how AI can drive tangible improvements in factory performance. But we’ve also seen the damage that comes from investing in the wrong solutions or falling for empty promises.

AI is only as good as the people managing it and the problems it’s solving. If you’re not clear on either, you’re setting yourself up for failure. Here’s what we’ve learned about successful AI implementation over many years:

• Start with Business Goals: Don’t implement AI for the sake of it. Identify specific pain points and target them with the right tools (which may or may not involve AI). Don’t fall for solutions that don’t align with your needs.
• Use Quality Data: AI is only as good as the quality of information given to it, without accurate, reliable data, AI can’t deliver. Ensure that your data is clean, structured and directly relevant to the business goals.
• Start Small, Scale Gradually: Begin with targeted projects that can deliver quick wins and build confidence, such as remotely monitoring production machines.
• Measure, Iterate, Improve: View AI implementation as an ongoing process, not a one-and-done project. Continuously monitor performance, learn from the results to refine your approach and integrate lessons learned into broader operational improvements.

AI Delivering Real Results

The manufacturers making real progress with AI aren’t chasing the latest trend, they’re taking a structured, problem-solving approach. Here are three examples of how our solutions are delivering tangible results:

1. Optimising Factory Performance
   Our Decision-Intelligence powered manufacturing solution integrates assets from across the factory. This provides critical insights into real-time machine performance, including key performance indicators (KPIs) and warnings of problems before they occur. For one client, this resulted in 15% higher machine uptime; 20% reduction in energy usage and 10% capacity gains.
2. Improved Quality Control
   By continuously monitoring machine performance and process data, another client reduced defects and rework to achieve 98% right-first-time production, alongside actionable insights for further process adjustments.
3. Streamlining Complex Product Configuration
   Our manufacturing solution is the glue that connects your data analytics, AI and in-house expertise with meaningful business outcomes. We can capture and automate the knowledge of specialist solution engineers required to deliver a full design specification and quote, without compromising on the customisation requirements.

A Strategy That Works

These aren’t theoretical case studies. Each one represents a real-world outcome from manufacturers who demanded measurable improvements. By insisting on transparency and proven results, you’ll be best positioned to unlock AI’s full potential.

Remember: those offering systems that work don’t rely on rhetoric – they show, don’t tell.

This article is part of our ongoing series, Reality Check: What AI Really Means for Manufacturing, designed to inform, inspire and help you implement AI effectively in your manufacturing operations.

Look out for our next article, where we break down how, despite the distractions of major AI announcements, the most successful manufacturers remain focused on robust, results-driven strategies.

And find out more about our Decision-Intelligence powered manufacturing solution HERE