The Dangers Lurking in Your Org Chart

Designed to Fail

Eleven men died because no one connected the dots. Not because the dots weren’t there.

The pressure readings, the broken BOP valve, the skipped cement test, the missing centralizers. Every signal was in the system. They died because the organizational machine running BP, Transocean, and Halliburton was built not to synthesize them.

This is the third and final piece in the Macondo series. Today we go to the root. It is the 170-year-old design that still dictates how most companies move information and authority, and why it fails catastrophically under pressure.

It starts with a 71L clerk in the U.S. Army in 1980s Germany.

Specialist Mendoza was a human carrier pigeon in an olive drab Volkswagen cargo van. Every day he carried our maintenance report to Regimental HQ in Nuremberg. The reports from our squadron and the other squadrons gave the commander an accurate picture of equipment readiness, just in case the Russian 47th Independent Tank Regiment decided to attack that night through the Hof Corridor.

Mendoza was solving a problem that has plagued organizations since the beginning of time. Creating a shared picture of the operating environment. Although it was less than 40 years ago, he delivered his report before the fax machine, before email, and decades before the cloud would make secure transmission immediate.

Our maintenance report was little different from the reports line superintendents on the New York and Erie Railroad filed in the 1850s. It listed broken equipment and gave the regimental staff a picture of how many vehicles could roll out the back gate to fight the Russians on any given day.

In September 1855, a railroad superintendent named Daniel McCallum sat in a smoky New York office and dictated what would become the world’s first organization chart. He wasn’t doing it for aesthetics. He had a problem. His railroad, the New York and Erie, was 500 miles long and bleeding money on a per-mile basis compared to shorter lines. His conclusion was sharp. The failure wasn’t the railroad. It was the absence of a system.

What McCallum invented wasn’t just an org chart. It was an organizational framework. There were clear divisions of responsibilities, delegated authority, daily data feedback, and multidisciplinary teams organized around the flow of value. Leaders weren’t at the top of his diagram. They were at the bottom, with reporting lines drawn like branches of a plant. That detail alone says something about how he saw the job.

McCallum’s design was built around a simple insight. The people closest to the work need real authority to act on what they know. His divisional superintendents had full accountability for their lines, full visibility into daily data, and the power to make decisions without waiting for New York to weigh in. The structure was designed to move information and authority in the same direction. He recognized the telegraph’s potential to shrink time and distance, and built his entire daily reporting system around it. For the first time, information could move toward authority at scale.

Then Frederick Winslow Taylor came along and reversed the logic. Scientific Management separated thinking from doing. Managers at the top planned. Workers executed. Functions organized vertically by specialty, each reporting up its own chain. The org chart stopped looking like a branching plant and started looking like a series of parallel silos. Coordination happened, in theory, at the top. It worked brilliantly for hierarchical manufacturing organizations for a time. Globalization and the explosion of information across the enterprise broke it.

Why does this matter? Why am I talking about an org chart for a long defunct railroad?

I am looking back in time because modern organizational design is, in part, responsible for the tragedy at Macondo I’ve written about over the past few weeks. The hierarchical organization design with no way to coordinate the flow of information across BP, Halliburton, and Transocean allowed the $128,000 cement bond log BP skipped and the $1 million-per-day rig costs to become a $65 billion liability.

Macondo shows exactly what a siloed structure produces under extreme financial and operational pressure. BP, Transocean, and Halliburton each owned a vertical column. BP controlled the well plan. Transocean owned the rig and the crew. Halliburton owned the cement job. Each organization reported up its own chain to its own executives, with its own incentives and its own definition of success. Nobody owned the integrated picture. Nobody owned the synthesis function.

The anomalous pressure readings from the negative pressure test on April 20th were observed by people on the rig floor. The mechanical problems with the BOP valve were documented. The authority to stop operations and resolve the ambiguity sat somewhere above them, distributed across three corporate hierarchies with no shared mechanism for synthesizing what they were collectively seeing. McCallum would have recognized the problem immediately. He’d designed his entire system to prevent collisions on a single track railroad line 155 years earlier.

McCallum understood in the 1850s what organizations are still struggling with today. Information has to move toward authority, and authority has to sit close to the work. When you invert that relationship and then multiply it across globally dispersed organizations operating at digital speed, complexity doesn’t grow linearly. It grows exponentially.

For the first time since McCallum picked up his pencil, we now have the technological tools to solve the very failure that killed 11 men on Deepwater Horizon. AI can aggregate signals across dispersed organizations, surface anomalies, and connect information streams that never talked to each other before. The organizations that embrace it intentionally, that build synthesis functions into their structure with clear governance and accountability, have a genuine opportunity to prevent the next Macondo. The ones that let AI propagate without that intentionality will find themselves with exponentially more data flowing across the same siloed hierarchies Taylor built. They won’t have solved the problem. They’ll have amplified it.

Had BP connected the dots between the skipped cement test, the broken BOP valve, the fewer than recommended centralizers, the anomalous negative pressure test, Macondo would not have spilled oil into the gulf and killed eleven men.

Specialist Mendoza was a node on a distributed organization carrying information to decision makers. Today, the data sitting in our data warehouses, ERP, CRM and decision support systems serve the same function. Now the key is to effectively use technology to connect the dots, to create clarity, to manage risk, improve decision making and help organizations act on information they already possess.

The real question isn’t whether you have the information. It is do you have the ability to act on it.

.