So why isn’t AI having the impact we expected?

‍ Activity without connection doesn’t create value.

‍There are two patterns I am seeing repeatedly in conversations with leaders and, increasingly, in the data coming through my AI Value Realisation Diagnostic.

‍ At one end are organisations that know AI matters, but genuinely don’t know where to start. The pace of change is overwhelming, the market is noisy and the advice is often contradictory. They may be experimenting with Copilot, Claude or ChatGPT, but beyond that there is no clear view of what they should be doing or what successful AI transformation actually requires.

‍Importantly, many are not asking questions about operating models, work redesign, job architecture, capability, governance or value realisation. They don’t recognise that these are the million-dollar questions. Much of the AI conversation still begins with technology: which tool, which platform, which use case, which agent?

‍ At the other end are organisations investing heavily and doing a lot. Pilots are multiplying, tools and agents are being deployed, people are being trained and AI initiatives are popping up everywhere. They are busy with AI, but that doesn't mean the organisation is transforming

‍ ‍

The gap between AI activity and enterprise value

‍McKinsey’s 2025 global AI research captures the gap well. 88% of organisations report regular AI use in at least one business function, yet only 39% report any enterprise-level EBIT impact from AI.

‍AI adoption is becoming mainstream. AI value realisation is not.

‍A successful pilot tells us that a particular technology can solve a particular problem. Training tells us people have been exposed to new skills. An AI policy tells us some governance is in place.

‍None of those things, individually, tells us whether the organisation is transforming.

‍That requires a much more connected set of decisions: understanding the problem worth solving, choosing where to invest, establishing what success looks like, redesigning work and workflows, considering what changes in the operating model, determining what happens to roles and capability, managing the change and ultimately measuring whether the promised value is realised.

‍Imagine an AI solution reduces the time required to complete an activity by 30%. Technically, that could be an excellent result. But what happens to the 30%?

‍Does the workflow change? Is the capacity removed, redeployed or reinvested? Are roles re-imagined based on the new realities? Are different capabilities now required? Could the capacity improve customer experience, increase revenue or address work that previously wasn't getting done? And six months later, can anyone demonstrate where the promised value actually landed?

‍If those decisions are never made, an organisation can create productivity without creating any genuine enterprise value.

‍The same problem appears in different forms. Training without redesigning work can leave people returning to fundamentally unchanged jobs with new skills. Automation without workforce planning creates capacity without deciding what to do with it. Pilots without clear success measures create demonstrations rather than investment decisions.

‍The individual activities may all be worthwhile. The value comes from connecting them.

‍IKEA provides a useful example. As its newly created chatbot increasingly handled routine customer enquiries, IKEA also reskilled thousands of customer-service employees into areas including remote interior design and sales. Its remote sales channel generated €1.3 billion in sales in FY22.

‍The significance isn't that the chatbot created €1.3 billion in revenue. It didn't. It is that IKEA connected automation to work redesign, workforce capability, customer experience and commercial opportunity. Instead of stopping at “the AI works”, the organisation considered what it could now do differently.

‍This is why successful AI transformation needs to be treated as a connected organisational system, not a collection of AI initiatives.

‍That thinking sits at the heart of The Klease Method™: understand where value can be created, design what needs to change across the organisation, and embed it in how the business actually operates.

‍The objective isn't to make AI transformation more complicated. It is to make it clearer what needs to happen, what needs to connect and what an organisation should do next.

‍ ‍

So, where should you start?

‍The next move will be different for every organisation. Some need to experiment and learn. Others have already proven the technology works, but are being held back by data, governance, adoption or an operating model that has not kept pace. In some cases, AI is already creating capacity, but work, roles and capabilities have not been redesigned to capture it. In others, there is simply no credible way of knowing whether the expected value has been realised.

‍ If you would like to understand where your organisation sits, I have developed a complimentary AI Value Realisation Diagnostic to help identify where you are strong, where the gaps are and where attention may be needed next.

‍Take the complimentary AI Value Realisation Diagnostic

‍For anyone interested in the IKEA example, you can read more about it here: ‍ ‍Read the Ikea Case Study Article

‍