The AI knowledge gap - and why it starts at the top

AI IS everywhere. You can't open a trade magazine, attend an industry event, or scroll through LinkedIn without someone telling you it's going to transform your world. And they're probably right. But here's what I keep seeing across the cleaning, FM, and environmental sectors: the people making decisions about AI often don't understand it well enough to make good ones.

I'm not trying to be provocative, although I think this probably is. I regularly meet senior leaders who tell me they "know enough about AI" and then reveal, within minutes, that theydon’t know the half of it. They've read the headlines about ChatGPT and autonomous robots. They might have played with a few free tools. But ask them how AI can specifically improve their tender process, their content output, their customer communications, or their competitive positioning, and they go really quiet.

This isn't intended to be a criticism. It's an observation about where we are as an industry. And it’s important, because when leadership doesn't understand AI's potential, it holds the whole business back.

The numbers don't lie

Research from Be the Business found that 42% of UK SMEs have no plans to implement AI tools in the next year. A YouGov survey put current AI adoption among SMEs at just 31%. Why? A 2025 report from ANS and techUK identified lack of expertise as the number one barrier to adoption, cited by 35% of businesses surveyed.

But the statistic that should really make you sit up: according to RAND Corporation research, more than 80% of AI projects don’t get beyond pilot stage. That's twice the failure rate of IT projects that don't involve AI.

Why is this? Because organisations rush in without understanding what they're doing. They buy the shiny tool, run a pilot, don't see results, and decide AI isn’t for them.

The real problem isn't technology

The barrier to AI adoption in our sector has nothing to do with the technology itself. Tools like ChatGPT, Claude, and Microsoft Copilot are accessible, affordable, and genuinely useful. You can get started for £20-30 a month. The technology isn't the problem.

The problem is knowledge. Or more specifically, the lack of it at the top.

When leaders don't understand AI, nothing happens. They're not sure where to start, so they don't start at all. And when someone does eventually push for action, it often takes the form of jumping straight to a specific tool or platform without laying any groundwork first.

I've seen this pattern quite a lot recently. A business invests in an AI tool because someone saw it at a show or read about it online. They roll it out without any proper training or understanding of how it fits into their existing workflows. When it underperforms, they blame the technology. But the tool wasn't the problem. The problem was bypassing setting up the foundations.

Foundations before fireworks

The key point is that you can't shortcut your way to AI success.

It's like buying the latest piece of cleaning equipment and giving it to an operative with no training. You blame the machine when the results are poor. The equipment isn't faulty; the implementation is.

Before any business in our sector invests seriously in AI, three foundations need to be in place:

First, get your information organised. AI can't help if your files are a mess. If your case studies live in seventeen different folders, your tender responses are scattered across multiple drives, and you can’t find the latest version of anything, AI will just amplify that chaos. Data organisation isn't exciting, but it must be done.

Second, train your people properly. Not a one-hour webinar and a handout - proper training on how to use AI tools effectively. Most people are getting perhaps 30% of what AI can deliver because they don't know how to prompt it properly. The tool is only as good as the person operating it. Research suggests that 81% of professionals think they can use AI, but only 12% have the skills to do so. That gap is where failure breeds.

Third, develop a clear AI policy. This covers data security, appropriate use, and governance. It's foundational and not an option. When you're handling client information, tender documents, and commercially sensitive data, you must know what can and can't go into these tools. Business-grade subscriptions to ChatGPT, Claude, or Copilot provide enterprise-level security where your data isn't used to train models. Free tools don't.

Get the foundations right, and the advanced applications will become possible. Skip the basics, and you'll join the 80% whose AI projects fail.

What AI actually offers our industry

So, what can AI realistically do for businesses across our sector? The answer depends on what you do, but the common thread is this: AI excels at tasks involving research, writing, analysis, and content creation. Every business in our industry does these things daily, whether you're cleaning buildings, manufacturing products, or selling them.

For service companies - contract cleaners, FM providers - the biggest wins tend to come from tender and proposal acceleration. The potential to cut response times by 50-70% is there when teams are properly trained for research, drafting, and quality checking. Beyond tenders, there's case study creation, client reporting, compliance documentation, and training content.

For manufacturers - AI transforms product documentation, technical specifications, marketing content, and regulatory materials. Market research that used to take days can be done in hours. Competitor analysis becomes something you can do regularly rather than once a year.

For distributors - it's product descriptions, customer communications, sales enablement materials, and internal training resources. The repetitive content tasks that eat into your team's time can be dramatically accelerated.

This is not about replacing people. Our industry already struggles with recruitment and retention. The point is to free your people from the grind of repetitive tasks so they can focus on the things that need human judgement: client relationships, quality, strategy, and creativity.

Start with marketing, then scale

If you're wondering where to begin, marketing is usually the answer. Not because it's the most important function, but because it's the safest place to prove the model.

Marketing tasks are visible and measurable. You can track time savings. You can compare output quality. You can demonstrate ROI in ways that operations or finance might find harder to quantify quickly. And the skills your marketing team develops with AI, transfer easily to other departments.

Once marketing has proven what's possible, the approach scales naturally. Sales teams can use the same techniques for proposal support and customer research. Operations can apply them to documentation and compliance. HR can accelerate recruitment materials and training content. Customer service can build response frameworks.

The companies ahead of the game aren't treating AI as a marketing tool or an IT project. They're building AI capability across the whole organisation, starting with leadership who understand the potential well enough to champion adoption.

AI is a company skill, not a department tool.

Making the business case

None of this happens without investment, and this requires a business case. Here's how to build one that your board will approve.

Start with time. Track how long a task takes before AI and after. If a tender response that took six hours now takes 90 minutes, that's four and a half hours saved. At a manager's hourly rate, that's potentially £150-200 per proposal. If you're responding to five tenders a month, the numbers get serious.

Then consider opportunity cost. Faster tender responses mean you can bid for more work. Win one additional piece of business because you could respond in time, and AI has paid for itself for the year.

Microsoft and WPI Strategy research suggests that AI adoption by UK SMEs could add   £78billion to the economy over the next decade. The opportunity is real, but it will only materialise for businesses that adopt properly.

Build measurement into your implementation from day one. You can't prove value without baseline data, and you can't secure continued investment without demonstrating returns.

Your first 90 days

If you're ready to move beyond basic awareness and into genuine AI adoption, here's a structure you can follow.

• Weeks 1-2: Audit your processes. Where are the time drains? What's repetitive? What requires research or content creation? Don't try to transform everything - identify two or three high-impact areas where AI could make a measurable difference
• Weeks 3-4: Set up your foundation. Choose one business-level AI tool and implement it properly. Train key team members with prompting skills and an understanding of capabilities and limitations. Draft your AI policy
• Month 2: Apply and learn. Pick your highest-impact area and embed AI systematically. Document what works and track time savings. Refine your approach. This is where you'll build confidence and start to gather evidence
• Month 3: Measure and plan. Calculate actual results and build the internal business case for wider rollout. Identify the next departments or functions to bring on board. Plan your company-wide capability development.

This shouldn't be a one-off project. It's how your company should work from now on. The businesses seeing real returns from AI treat it this way.

The window won't stay open forever

First mover advantage in our sector is still available. While 42% of SMEs have no plans to adopt AI, and genuine expertise remains thin on the ground, there's time to get ahead.

But that window is closing. The businesses that invest in genuine AI understanding now - not just buying tools but building capability will have a two-to-three-year head start on competitors still trying to work it out.

The question isn't whether AI will transform our industry. It's whether your business will be ahead of the curve or behind it.

The answer starts at the top.

Ceris Burns is the founder of CBAi providing specialist AI training and consultancy for the cleaning, FM, and environmental sectors. A Fellow of the Chartered Institute of Marketing and Chartered Environmental Cleaner with over 20 years' experience in essential industries, she helps organisations move from AI awareness to confident implementation.

Find out more at cbai.co.uk