The Decision-Making System You Never Designed
Why decision wiring is the most visible form of leadership debt, and why frameworks alone will not fix it.
Every company has a decision-making system.
Most of them were never designed. They were inherited from the version of the company that no longer exists.
The wiring you laid down at one stage of the company quietly outlasts the stage it was built for. The approval gates set up when trust was being built remain in place long after the trust was earned. The senior leader who used to be the right person to consult is still being consulted by habit. The meetings that exist to confirm decisions already made are inherited from a version of the company you have moved past. None of this gets reviewed. It compounds.
This is decision wiring. And it is the most visible form of leadership debt.
Where the wiring fails
In a recent Harvard Business Review piece, Jonathan Rosenthal and Neal Zuckerman argue that consensus-driven decision-making is a relic. It made sense in slower, more dispersed times. It does not make sense in an environment where competitors move at AI speed and information is already stale by the time it has travelled up through three layers of curation.
They are right about the symptom. They are partial about the cause.
Decisions in scaling companies do not fail because the team is committed to consensus as a value. They fail because the wiring beneath the decisions has not been touched in years. The wiring produces three predictable failure modes.
The first is the slow-walk. A decision is technically open, technically being considered, technically progressing. In practice it is being held in suspension by a senior leader who has not said no but has not said yes. Other work cannot move until this decision lands. The slow-walk is not refusal. It is procrastination disguised as diligence.
The second is the ratification meeting. A decision has already been made informally, usually by the loudest voice or the longest-tenured leader, and the meeting exists to confirm it. Everyone in the room knows this. Nobody names it. The meeting takes 60 minutes to do what could have been a one-line message, and the team leaves believing they have decided together when they have actually only ratified.
The third is the pocket veto. A senior leader withholds agreement quietly. They do not disagree openly. They simply do not commit. The decision sits, suffocated, until someone else takes responsibility for it or it expires. The pocket veto is the most expensive of the three because it costs trust as well as time.
These are not behavioural problems. They are wiring problems. The same people, dropped into a different decision-making system, would behave differently.
Frameworks are not the answer
The management literature has been producing decision-rights frameworks for decades. RACI. DACI. OVIS. The latest of these names an Owner, a Veto holder, an Influencer, and a Supporter for each decision.
Source: HBR, April 2026.
These frameworks are useful. They name accountability where it would otherwise diffuse. As Rosenthal and Zuckerman note, “when accountability is diffused, speed is impossible.”
But naming the Owner does not build the Owner's judgement. Naming the Veto holder does not give the team the trust to use it well. Naming who Influences and who Supports does not change the underlying culture of slow-walking, ratifying, and pocket-vetoing that the team has practised for years.
The framework is the easy part. The conditions underneath it are the work.
This is where the consulting literature consistently underdelivers. The frameworks proliferate. The diagrams get pinned to the Notion page. Then the team goes back to the same patterns it had before, because the system that produced those patterns was never coached.
Decision wiring is not redesigned by drawing a new chart. It is redesigned by the team doing the actual work of looking at how they decide, naming what is not working, and rebuilding the conditions for honest disagreement, real commitment, and clean accountability.
That work is not theoretical. It is integration. And integration is what most decision-rights initiatives skip.
The capability AI is asking us to develop
Step back from the wiring for a moment. AI raises a different question, and it is the one most leadership development conversations are missing.
The question is not whether your team can use AI. They can. They are. The question is what capability humans need to build now that AI handles so much of what used to fill our days.
The honest answer, from years of working with leaders and teams, is this. The most important capability for the next decade is not technical literacy or AI fluency. It is the capacity to think critically, judge well, and decide together.
Critical thinking is the capacity to question what has been decided and what has been assumed, even when the assumption is your own. AI cannot do this for you. It will confidently produce an answer that sounds reasonable. The work is yours to know whether the answer is right.
Judgement is the capacity to decide on incomplete information. It is built through structured exposure to consequential decisions, not through grunt work or osmosis. AI is now doing the grunt work. The apprenticeship pathway that used to build judgement quietly is breaking, and most companies have not noticed.
Decision-making processes are the structures that distribute these capacities across the team rather than concentrating them in one person. They are not built by a framework. They are built by the team doing the work of deciding together, and noticing what is helping and what is in the way.
These three are not separate skills. They are a connected capability, and they cannot be taught. They have to be surfaced.
The three questions worth sitting with
In team coaching, the work of redesigning decision wiring does not start with a framework. It starts with three questions a team sits with together.
1. How have we made decisions until now?
2. How has that served us, and where has it slowed us down?
3. How do we need to decide, as a team, in a world where AI is faster than we are?
These are not questions to answer in a 30-minute meeting. They are questions that take a team several conversations to answer honestly. The first one is harder than it sounds, because most teams have never named their actual decision pattern out loud. The second one requires the courage to look at cost. The third one requires the imagination to think about decisions you have not yet had to make.
A team that does this work will not need OVIS. They will land on something better, because they will have built it themselves.
The work is the system
Beneath every performance problem is a system problem. AI does not change this. It applies pressure to it, and decision wiring is where the pressure shows up first.
You can buy a faster AI tool. You can adopt a sharper decision-rights framework. You can hire a senior leader you hope will fix the dynamics. None of those moves will work if the underlying decision-making system has not been touched.
The shift from operator to architect is not about delegating better. It is about building a system that decides without you in the room, and trusting that system enough to let it.
The leadership layer beneath you is doing the actual work of running the company. Whether they can decide well together is the single biggest determinant of whether the next phase of growth happens or stalls. That capability is built deliberately or not at all.
Comments