A question that keeps returning
After twenty-five years working across different industries, I keep returning to a question: why do familiar organizational problems survive so many serious efforts to solve them?
The businesses can be very different. Yet I have encountered similar difficulties, often followed by similar responses. People understand something important, but that understanding struggles to influence what happens beyond their immediate area. A problem involves several functions, each contributing useful work, while the difficulty they share remains unresolved.
I have worked inside these situations, helping to shape responses through data, governance and technology. Over time, they have made me question what we expect those responses to accomplish, and what else needs to change for them to succeed.
AI gives that question a new urgency. It can make each part of an organization more capable while increasing the consequences of what remains disconnected between them.
That is why I have built my work around Organizational Intelligence.
Joining an existing conversation
The term itself has a substantial history. Harold Wilensky published Organizational Intelligence: Knowledge and Policy in Government and Industry in 1967. Related traditions have explored how organizations learn and interpret their circumstances. Chris Argyris and Donald Schön examined the difference between correcting an error and questioning the assumptions that produced it. Karl Weick explored how people make sense of situations and how those interpretations shape organizational behaviour.
I am joining an existing conversation. What brings me to it is the persistence of problems I have encountered in practice, including in work to which I have contributed.
Where it becomes visible
One place where these difficulties become visible is the relationship between data teams and the wider business.
I have led data and governance work for much of my career. People generally want their work to make a difference. Data, analytics and digital teams bring expertise that can help an organization see and do things it could not before. Colleagues elsewhere bring an understanding of customers and daily operations that is just as necessary.
Yet it is surprisingly easy for us to grow apart while trying to contribute. We work to different priorities, with different measures of success. A technically sound delivery can represent real progress for one team while leaving another unsure how it helps with the difficulty they face. Both may have done their work conscientiously. Making our contributions useful together still requires attention that neither mandate fully accounts for.
Data quality makes this particularly visible. Business colleagues are asked to participate, sometimes in language that makes discussing their own information sound as though it requires a master’s degree from MIT. Someone who understands exactly why an order cannot be fulfilled can suddenly feel out of their depth when the same problem is presented as a data issue. That operational understanding is exactly what the work needs, and those of us leading data work, myself included, do not always make the connection easy to see.
The same pattern appears well beyond data. Procurement negotiates a cheaper packaging supplier and meets its savings target. Months later, the warehouse is repacking damaged goods and customer service is handling the complaints. Each function did what it was measured on. Nobody was measured on the whole.
Over time, I have become more attentive to what happens between these contributions: how people come to understand the problem together, and what allows their combined knowledge to change the way the organization works.
I see a similar tension in our reliance on particular people. Every organization benefits from colleagues who remember the history and know how to navigate the exceptions. Much of what works depends on their generosity and resourcefulness. Michael Polanyi called this tacit knowledge: we can know more than we can tell. The difficulty emerges when we repeatedly need the same person to make something work, without finding a way for their understanding to become more widely usable. We may only discover how much depended on them when they change roles or leave.
This complicates the idea that an organization “knows” something. Knowledge can exist inside it without reaching the place where it would be useful, or without carrying enough context to influence a decision.
Even when understanding is shared, action can remain difficult. People face commitments they cannot easily abandon. Incentives pull in different directions. A change that benefits the organization may impose a real cost on one team. These tensions require judgment and negotiation, and more information does not settle them.
We already have disciplines that address these conditions. Data governance makes information more trustworthy and usable. Knowledge management helps experience remain available beyond the people who acquired it. Organizational design shapes how responsibilities and relationships support the work. Decision-making practices strengthen how choices are framed and implemented. These fields overlap, and capable practitioners work across their boundaries. Organizational Intelligence draws on all of them.
What I mean by Organizational Intelligence
The value of Organizational Intelligence, for me, is in keeping one question in view throughout the work: is the organization becoming more capable of using what it knows and learning from what it does?
By Organizational Intelligence, I mean an organization’s ability to act on what it knows, and to keep getting better at it. In practice, that means developing an informed understanding of its circumstances and learning from experience to act more effectively.
That understanding does not require everyone to agree. It requires enough shared context for people to examine their differences and coordinate action. Learning, in turn, has to influence subsequent behaviour. A lesson can be carefully documented and still leave the conditions that produced it untouched.
This is where the focus extends beyond improving a decision or redesigning a management process. What matters is whether those improvements accumulate into a capability the organization can sustain. Organizational capability is what continues to shape coordinated action as individuals change.
Why now
AI is expanding what people and functions can accomplish, and those gains enter organizations with existing dependencies and unresolved disagreements.
A 2023 field experiment with consultants at Boston Consulting Group, “Navigating the Jagged Technological Frontier”, found substantial improvements on some knowledge tasks alongside poorer performance on others. Where and how AI helps matters. But even where it works well, we still need to examine what happens across the organization.
Imagine Commercial developing proposals faster, supported by stronger analysis, while Operations improves planning with equally capable tools. If the two functions use different assumptions about demand or capacity, each can improve its own work while the business takes on commitments it struggles to fulfil.
The issue may have existed for years. Greater capability makes its consequences harder to contain. A difference in interpretation can shape a larger volume of work before it is noticed. An assumption never examined across functions can become embedded in automated action. The cost of resolving the discrepancy grows as more activity depends on it.
AI can also help bridge these gaps, making relevant experience easier to find and conflicting interpretations easier to compare. That is part of its promise. Realizing it still depends on how the organization works: whether people have room to challenge an assumption and ways to resolve the implications, and whether what they learn can influence action, including when doing so is inconvenient.
Where to begin
Organizational Intelligence keeps that work in view. It is not a tool to purchase, a new department to establish or another framework to implement. Frameworks can help structure an intervention, but the capability has to develop through the organization’s everyday practices.
This is the perspective I am building CERNYVA around. Its value will have to be demonstrated through the work, and through what clients become able to sustain for themselves.
I return to these questions because I have seen how much expertise and effort organizations already contain, and how difficult it can be to make those contributions useful together.
As we invest in AI, that deserves as much attention as the capabilities we are adding. There is knowledge people have been trying to put to work for years. Understanding what has prevented them from doing so is where to begin.
