For years I did something that made me unpopular in meetings.
I asked the same question of three different leaders in the same company. They were never in the same room, and the conversations were spread across a week. I asked a CFO on Monday, a COO on Wednesday, and a head of customer success on Friday.
The question was always some version of the same thing: how does work actually move through this business, from the moment a customer commits to the moment we deliver value?
Over time, I posed that question in dozens of companies that were public, private, post acquisition, growth stage, and mature.
They represented different industries and sizes. The pattern held almost every time. I got three different answers.
None of the answers were contradictory in the dramatic sense. Each one was reasonable, and every leader had thought it through while describing real work that real people did. Still, the three descriptions did not fit together.
The handoffs the CFO described were not the handoffs the COO saw. The decision points the COO named were not the ones customer success lived with. The bottleneck each leader identified sat downstream of someone else’s bottleneck they never mentioned.
Put the three accounts side by side and you did not get one picture of the business. You got three overlapping ones with no agreement on which was true.
That was never a sign of bad leadership. These were capable, experienced executives who knew their functions cold. Their knowledge was never the problem. The organization underneath them had become illegible, and each had adapted to it in a slightly different way.
When Organizations Became Illegible
Illegibility never arrived in one moment. It accumulated, usually after a period of success and company growth.
- New offerings and teams were added while new leaders brought new ways of working.
- Acquisitions stitched in processes that almost matched the old ones but not quite.
- Reorganizations moved people across functions and the institutional memory of how things connected walked out with them.
Each change was small. Each change was justified. And each change shifted the actual flow of work a few degrees away from anything anyone had documented or said out loud.
After enough of those small shifts, the company was still operating, often well, but on a foundation nobody had looked at in years.
Work got done through workarounds that had hardened into rituals. Tribal knowledge filled the gaps documentation no longer covered. Different teams used the same words to mean different things and never realized it. Leaders had learned to operate without a shared map, because the people around them had learned to compensate for not having one.
From the outside the company looked coherent. From the inside it was held together by individual heroics and pattern recognition built up over years.
That was the company where my three leader question produced three different answers. This happened because there was no shared picture left to be right about, not because anyone was wrong.
AI Raised the Stakes
Organizational illegibility was tolerable for a long time because most enterprise technology adapted to it. ERP systems accommodated local variations, while CRM tools absorbed the exceptions. People filled the gaps the systems could not. Although the illegibility created friction, companies managed it and kept the business running.
AI changed the math. AI does not adapt to illegibility.
Applying it to an unclear system produces more than a slightly worse result compared with a clear one. It creates a different kind of failure, even when the models work and the outputs look right. However, the organization cannot absorb what the AI is producing because there is no shared picture for the output to attach to.
Three leaders read the same AI recommendation three different ways, because each is running it through a different mental model of how the business works. What had been a quiet inefficiency turned into a multiplier on AI investment. Every initiative got refracted through whichever leader’s mental model dominated the room that day.
Although the technology produced clarity, the organization absorbed it as ambiguity. That is what most failed AI pilots were actually showing. The technology was ready, but the organization underneath was never legible enough to use it well.
Documentation Made It Worse
When leaders sensed the illegibility, the instinct was to commission a project, hire a firm, map the processes, build a playbook, and roll it out.
That rarely worked.
It rarely worked because the deliverable became the artifact instead of the understanding. The firm produced a process map, presented it to the company, and received everyone’s approval. Then the map went on a server. Six months later the company had reorganized, two acquisitions had closed, and the map described a business that no longer existed.
The illegibility came back.
The map stayed on the server. Documentation does not produce organizational understanding. It creates a static description that may be useful for a few weeks, while the system itself keeps moving.
What produced understanding was never a deliverable. It was a practice.
The Practice That Closed the Gap
This is the practice I ran as a CIO, and the one I would push any CIO to adopt. It is not glamorous. It does not need a budget. It needs discipline and a willingness to be slightly annoying in meetings.
Ask the question. Then ask it again. Then ask it again.
I picked one or two macro processes:
- Order to cash.
- Idea to launch.
- Customer onboarding.
- Renewal.
The ones that mattered most to value creation.
Then over the course of a quarter I asked the same simple question about that process in every leadership conversation I was already sitting in. It was not a project or a formal interview. It was a habit.
- How does this actually work?
- Where does the handoff happen?
- Who decides?
- What signals the next step?
- Where does it stall?
I asked the CFO in a budget review, the COO in an operations sync, the head of sales on a forecast call, and the head of customer success in a renewal discussion.
I asked the same question lightly, in passing, over and over.
Then I paid attention to the gaps. The gaps between the answers were the real data. The gaps between answers were where the organization was illegible.
- They told me where decisions were contested and nobody had named the contest.
- Where handoffs were informal and had never been agreed to.
- Where the same word meant different things in different functions.
- They also revealed where an AI investment was going to fail before I made it.
Most CIOs are trained to fix things. This practice was not about fixing. It was about seeing. The fixing came after, and it came faster, because by the time I was ready to invest in process clarity I already knew where the illegibility lived. I had a map of the map’s holes.
The leaders I watched use this well ended up doing something rare in a C suite. They became the person who could describe the business across functions, because they had heard every function describe it and had noticed where the descriptions diverged. That made them hard to do without in AI conversations, transformation conversations, and board conversations.
Their value came from having more accuracy, not more authority.
The Hard Part
It was slower than commissioning a consultant. It did not produce a slide or give a board the satisfaction of a deliverable. For the first quarter it looked like nothing was happening.
What was happening is that I was becoming the only person in the company who actually understood how the company worked.
That was not a credential. It was leverage. It was the position from which every AI investment decision, every modernization choice, every transformation conversation got better. That improvement came from doing the work of seeing, not from being smarter.
Most organizations do not understand themselves. The ones that came to understand themselves never got there through documentation. They got there through leaders who made it a practice to ask the same question in enough rooms, with enough patience, to finally hear what the organization had been saying the whole time.
Three leaders, three answers. The gap between them is where the work is.
