- One reason why AI isn’t paying off yet is that companies buy licences without investing in how their teams work.
- The ICT and digital transformations worked because companies invested: they defined a bold future state, assigned ownership and set a realistic budget.
- AI needs the same. Fund pilots, training and new ways of working, not just licences, and invest where AI makes your business unique.
In this article
AI is everywhere, and yet, in most companies, it is nowhere. Every board talks about it and every employee has a chatbot. Few businesses have changed how they actually create value.
I see today’s excitement as the herald of what will, and should, happen next. Having lived through both the ICT transformation and the digital transformation, I notice one important thing missing from most AI conversations: investment.
What the digital transformation taught us about AI
In the ICT transformation (yes, I am either that old, or it was not that long ago), introducing Voice over IP meant putting money on the table. Companies ran pilots, trained users on the new capabilities and fixed the early problems.
The digital transformation was no different. A digital customer platform required investment, new processes and a rethink of how people work, so the customer experience became consistent. Employees did not bring their own phones to the office, and a new website alone did not give anyone a better customer experience.
That is why it surprises me to see businesses laying off people on the assumption that AI will fill the gap. And it does not surprise me that some are now hiring people back. Forrester found that 55% of employers who cut staff citing AI regret it, and Gartner expects half of companies that cut customer service staff for AI to rehire by 2027 (People Matters, July 2026). The reason is simple: investment comes first; new revenue and efficiency are the results.
“
Investment comes first; new revenue and efficiency are the results.
Better performance is not yet transformation
As Prof. Didier Bonnet of IMD rightly puts it, the current wave of generative AI mostly helps the workforce perform better. That is the first of three levels.
Public tools make work more efficient when they work with public data. Putting business-critical data into public platforms is something European companies must handle with care.
Inside a business, tools like Copilot stay in the space of the individual user. They summarise the meeting I missed and help me keep my inbox under control. That optimises my workday. It is not the leap forward for the company.
What AI transformation really means
AI transformation moves value creation from a digital state to an AI-embedded one. That is fundamentally different from making things faster or cheaper.
The key question is a company’s intellectual property: its products and services. Delivery and operations follow these priorities, not the other way around. Building something new will always be inefficient at first, and the running business has to carry the investment. So it cannot be neglected either.
Three moves that worked before, and will work again
How did we master the digital transformation? We defined a bold future state, assigned ownership and set a realistic budget. AI is no different, and it must be a management topic, not an afterthought anywhere, including manufacturing, healthcare and other traditional industries.
1. Define a bold future state
Try to envision the unthinkable. Sketching an ambitious, even utopian, picture of your business opens up new options and opportunities. Every industry has things that could set one company apart but were never worth doing by hand. AI can finally make them worthwhile. A clear, prescriptive future state is the critical starting point.
2. Assign ownership
Roles like the Chief Digital Officer or Chief Digital Transformation Officer were transformational roles. Their job was to create awareness, back the first initiatives and grow the organisation from first solutions to a power-user community across business units, and finally to knowledge shared across the whole workforce.
The transformation was complete when business units already thought in digital solutions, and IT had the capabilities and environments to connect and support them. Today’s Chief Data Officers and Chief AI Officers play the same role: building awareness and stretching the organisation to build its muscles, on the business side and in IT.
3. Set a realistic budget
The knowledge has to be built within the workforce; your people are the experts. But they need support to change routines that have run for years. That means breaking up old approaches while keeping the ship running. A pure revolution does not work: you do not put the old furniture out on the street before the new one has arrived. Budget for that change (pilots, training, new ways of working), not just for licences.
Budget for the change, not just for licences.
Why AI adoption stalls, and what protects your uniqueness
It is no surprise that only about 3% of the roughly 450 million commercial Microsoft 365 users pay for Copilot, around 15 million seats (The Register via TechRadar, February 2026). We saw the same with Office and most other tools handed to users: something added on top does not naturally find its way into established routines. Even less so when it lacks the business context to get a job done, rather than just making the result a bit shinier.
Two things have to come first:
- Business understanding. The business must understand how AI becomes part of its value creation.
- A secure data and AI setup. Company information must be available to every user, within their permissions. This is the only way to protect what makes you unique, and ultimately your IP.
If everyone uses the same tool to answer the same questions and build the same presentations, companies will start to look alike. Uploading all your data into a public AI environment will hardly give you, say, a proper sales calculator that applies your pricing, regional and marketing rules. And by the time those tools are sophisticated enough to do it, your uniqueness is gone, because everyone can do the same.
Take the first step: invest
Unlike the industrial age, when building a business needed heavy upfront investment, AI makes it cheap to start a business and get relatively far. That cuts both ways. Whatever is easy to automate will be simplified, commoditised and lose its relevance.
Companies that want to stay relevant must invest in AI where it creates unique selling propositions, and with them IP that builds shareholder value.
So take the first step: invest.
Frequently asked questions
Why isn’t AI paying off in most companies?
Because most AI so far helps individuals work a little faster, while the business itself has not changed. Companies bought licences but did not invest in the change: a clear future state, ownership and a budget for pilots, training and new ways of working.
How is AI transformation different from digital transformation?
The playbook is the same: invest, assign ownership and change how people work. The difference is the goal. AI transformation moves value creation from a digital state to an AI-embedded one, so it reaches a company’s products, services and intellectual property, not only its processes.
What should a 2027 AI budget cover besides licences?
Pilots, training, new ways of working and the secure data and AI setup that lets every user work with company information within their permissions. Licences alone rarely change established routines.
Who should own AI transformation?
A role with a transformation mandate, as Chief Digital Officers had in the digital transformation. Today that is usually a Chief Data Officer or Chief AI Officer, working with both the business and IT.
What are the three levels of AI transformation?
Following Prof. Didier Bonnet of IMD: equipping the workforce, rewiring the operating model, and innovating the core. Only the third changes how the company creates value.
Free Webinar
Time to Make Your AI Pay Off
get ready for the EU AI Act and plan your 2027 AI budget.
Thursday, 26 November, 2026
14:00-14:45 CET • 45 minutes, in English, with Ivan Jelic
Want to know where AI can create real value in your business? We help you move from first pilots to AI embedded in your core.
The post Why AI isn’t paying off yet, and what to invest in first appeared first on Joyful Craftsmen.