You can’t escape it. AI is the next frontier in business transformation. Faster processes, better outcomes, smarter decisions. Yet in practice for many companies, the results are mixed. Our latest Work That Works report, based on insights from 2,000+ small business leaders and employees, uncovered a critical warning: in organisations where AI is poorly implemented, employee productivity can decline by up to 50%.
This isn’t just a case of missed potential. It’s a systemic failure.
What’s going wrong? We’re obsessing over tools while ignoring the people expected to use them.
Strategy Is Missing, Not Software
The 2020s have been dismal for productivity growth. Small and mid-sized enterprises (SMEs) are being told AI is the silver bullet, but without guidance, many are falling behind. In the UK, 99% of businesses are SMEs and our data shows that AI is now their number one challenge. It’s ranked above hiring or inflation concerns.
There’s high-level enthusiasm: 50% of business leaders say AI is already boosting productivity. But only 28% of employees use it weekly, and just 9% use it daily. That adoption cliff isn’t a minor issue, it’s a signal that top-down transformation is failing to connect with frontline workflows.
AI Fails Without Cultural Integration
Inconsistent or poorly supported AI adoption correlates with sharp drops in output. But in companies where AI is embedded intentionally with clear use cases, ongoing training and aligned expectations, performance significantly improves.
Crucially, employees don’t see AI as a primary productivity driver. It ranked 20th in our research. Meanwhile, leaders are twice as likely to view tech innovation as the key to output. That disconnect isn’t just philosophical; it’s operational.
You can’t architect digital transformation from the executive suite alone. Real value emerges when AI is integrated into the day-to-day habits of teams across the business.
SMEs Are Struggling. But They Stand to Gain the Most
Smaller businesses are navigating this shift with fewer resources. Only half report keeping pace with tech change over the past year. The challenges are steep: firms with fewer than 50 staff are 50% less likely to have deployed AI and 30% are operating in survival mode.
Yet the upside is real. Employees who use AI regularly rate their own productivity 17% higher. They’re more likely to say they’re making an impact at work and in companies where automation is visible and useful – not hidden or confusing – productivity is twice as high.
The message is clear: AI works, but only when the implementation gap is closed.
What we’ve learned at Employment Hero
At Employment Hero, we treat AI not as a side initiative, but as an org-wide transformation. We’ve iterated, experimented and learned. Here’s what’s worked for us:
- AI as a Strategic Lens, Not a Side Project
Every quarter, we align the company around a core focus and this quarter, it’s AI. That means every team, from engineering to marketing, is exploring how AI can improve workflows and outcomes.
We’re not pushing AI from the top down. We’re igniting grassroots curiosity and experimentation.
- Build Trust Before You Build Tech Stacks
We didn’t drop AI on people overnight. We started with conversations early, open and continuous. Our onboarding embeds an AI-first mindset. Mandatory training isn’t “How to use ChatGPT 101” – it’s about real scenarios where AI reduces cognitive load, unlocks creativity and enhances human value.
As confidence grows, so does adoption.
- AI as a Creative Companion, Not Just a Tool
We gave people space to explore. Dedicated Slack channels for AI experimentation have helped teams – from engineers to designers to recruiters – share what they’re testing and learning. This has been an amazing opportunity to collaborate and learn between traditionally siloed teams.
We highlight smart ideas and reward creative thinking, not just perfect outputs. That’s helped build a culture where people want to use AI because it’s genuinely helpful – not because they’ve been told to.
Adoption Metrics Aren’t Enough
Tech leaders often measure success by deployment milestones, usage dashboards, or cost savings. But those are lagging indicators. The real question is: do your people understand how AI fits into their work – and do they trust it to help?
If the answer is no, the best model in the world won’t deliver ROI.
What’s Next? Make AI Human-Centric
AI isn’t a fix-all. It’s a force multiplier. For teams that are empowered, trained and aligned. Without that, your investment is a sunk cost.
The future of AI in business won’t be shaped solely by technical breakthroughs. It’ll be shaped by how well we bridge the chasm between innovation and inclusion.
Because ultimately, AI the future of work isn’t being lost to AI, it’s being reshaped by it and core to that is how it’s implemented.
