There are three sentences you probably know. You rarely hear them in the steering committee and constantly in the corridor. The first: “Technically everything works. It is just that nobody logs in.” The second: “Everyone wants it, but we have been building for eighteen months now.” And the third, the quietest one: “The team is strong, the platform is up. But when the CFO asks what it has delivered, the room goes silent.”
Three companies, three different problems? Not at all. That is the same problem three times over, in three disguises.
I draw the same picture in nearly every workshop: three overlapping circles. Business, people, technology. Value only appears in the middle, where all three intersect. It looks like a consultant’s slide, granted. But the picture has one property that lifts it out of slide hell: it tells you in advance how you are going to fail. Every missing corner produces its own recognisable variety of failure. Precisely the three sentences above.
So that this does not stay abstract, let us go out to eat.
Menu, brigade, kitchen
A good restaurant needs three things, and none of them is negotiable.
It needs a menu. Someone has decided what reaches the table: which dishes, for which guests, at what price. The menu is not decoration, it is the commercial decision of the house. In a company that is the business circle: it defines where the value sits. Which decisions, processes and products are meant to get better through data and AI, which value pools are worth going after, which use case has an owner with a name.
It needs a brigade. People who cook, plate and season every evening, and do it in a way that leaves them willing to come back tomorrow. That is the people circle: roles and responsibilities, skills, a governance model that works as collaboration rather than as a catalogue of prohibitions, and a culture in which numbers beat opinions. People carry adoption, every single day.
And it needs a kitchen. Stoves, refrigeration, equipment, routines: what turns twenty covers into two hundred without the quality collapsing. That is the technology circle: a toolbox of platform, services and standards, built for scalability, security and usability. The kitchen is the foundation of the evening. But nobody books a table because of the stove.
The gaps have names
Now take away one circle at a time and watch what happens. That is the real value of the picture: the failures are not diffuse, they are diagnosable.
Menu and brigade without a kitchen: the delivery problem. Ambitious dishes, motivated cooks, four hobs. Every order becomes an act of strength, waiting times grow, and eventually the head chef starts sentences with “we cannot do it like that”. In a company: the use cases are agreed, the business is fired up, but the data foundation is missing. Every dashboard is a handstand, every model a one-off. “We have been building for eighteen months now.”
Menu and kitchen without a brigade: the adoption problem. The new combi steamer is the most expensive shelf in town, because nobody learned to cook with it and nobody was asked whether they wanted to. In a company that is the platform with five active users, the forecasting model whose list nobody opens, the chatbot everyone tried exactly once. “Nobody logs in.”
Brigade and kitchen without a menu: no ROI. Everyone is busy, everything runs, the kitchen gleams. It is just that nobody decided what the guests are supposed to order and what it has to bring in. What gets cooked is whatever appeals. In a company that is the technically impressive data and AI landscape without prioritised use cases: plenty of operations, no effect. “When the CFO asks, it goes quiet.”
Why “enable” does not mean “second fiddle”
Business leads, people and technology enable: the sentence is often misread, as though it described one boss and two suppliers. It means something else. Leading means setting direction and standard. Enabling means making them reachable. Without the two enabling circles, leadership is a wish list; without the leading circle, the other two are an end in themselves.
One detail surprises many people: the functional architecture of the data and AI ecosystem belongs on the business side too. Not the question of which database it will be, but very much the question of which capabilities the house needs. What must our ecosystem be able to do for the menu to be cookable? Delegate that decision entirely to engineering and you get an architecture optimised for elegance rather than for your own menu.
And the order? The circles are not a relay. The phased model of “platform first, then training, then value” sounds sensible and reliably produces two of the three gaps above. Adoption does not happen after the building, it happens during it: when the future users are standing at the stove, when their objections make it into the recipe, and when the first use case lands in their working day instead of in the demo folder.
And now the asterisks
First: the circles are not equally hard, and their difficulty moves. Fifteen years ago technology was the narrowest circle; delivery problems were the norm. Today the level of technology that satisfies most companies is simply solved and available under contract. That circle fills almost by itself. So the failures migrate to where you cannot buy anything: to adoption and to ROI. If you had to bet on what your next initiative will founder on, do not bet on the technology.
Second: AI tightens the people circle rather than relieving it. Adoption used to mean using a tool. With agentic AI, adoption means handing over a decision. Anyone asked to leave invoice checking to an agent is not giving up three clicks but a piece of judgement, and will only do it if the guardrails, the traceability and the trust are there. A training day does not tick that circle off; you find that out in operations at the latest.
Third: the diagram is a diagnosis, not a plan. Three full circles are necessary, not sufficient. Where to start, how momentum builds, how a first proof turns into a portfolio: that is what the other tools in this series are for, from the potential map to the proof of value. And in a fifty-person company, two of the circles are often the same person. That makes a lot of things easier and one thing more dangerous — you do not notice when one of the two hats stays unworn.
Back to the dining room
In the end the Venn diagram is a simple promise: if you know which corner is missing, you know what to do. The sentence from the corridor is not a mood, it is a finding. “Nobody logs in” does not mean “our people are refusing”, it means “we forgot the brigade”. “We have been building for eighteen months” does not mean “IT is too slow”, it means “we gave the menu no kitchen”. And the silence around the CFO is not a verdict on data and AI, it is the receipt for the fact that nobody ever decided what belongs on the menu.
So a good first step is not a tool selection but an honest look at all three circles at once — an assessment that puts value definition, adoption and foundation on the same table in a single session.
The restaurant you like eating at has all three. You notice it because you hear none of the three sentences, only the sound of a full dining room.
It only fills up in the middle.