Michael Dell Put a Glass Wall Between Himself and His COO. Here's Why It Worked
The COO is the least studied seat in the C-suite. Nate Bennett has spent twenty years fixing that. He co-wrote Riding Shotgun: The Role of the COO with Stephen Miles, first published in 2006 and now in its third edition from Stanford Business Books, and he runs the executive MBA program at Georgia State University's Robinson College of Business.
On this episode of the Strategy and Leadership Podcast we talked about what makes the role work, what makes it fail, and why the average COO now lasts about half as long as everyone else at the table.
Stop asking what makes a great COO
Nate reframed the question before answering it. The useful question is what makes a great CEO and COO pair. Both executives have to be effective, and that depends on a relationship built on trust that the rest of the organization understands.
His test is a family one. When mom says no, the kids go ask dad. People in organizations do the same thing. If the COO says no, they need to know the CEO will back them. There can be no daylight between the two, and everyone in the building has to know it.
The founder trap
Sharing power is hard for any CEO, and hardest for founders. The business is personal. Handing meaningful authority to someone else takes humility that not every founder has.
Nate described a pattern he's seen repeatedly. The board or investors see the founder losing control and insist on a COO. The founder agrees, picks someone, and then never empowers them. The COO gets blamed for the lack of results and exits. The actual failure was the CEO not being ready to admit they needed help.
One COO he interviewed for the book put it bluntly: anyone who takes a COO role from a founder is nuts. Not because the relationship was bad. Because of how hard it was to help the CEO get comfortable not controlling everything.
When to hire a COO
Two factors decide the timing. The first is the business itself. Operationally intensive companies, whether because of production or supply chain complexity, tend to add a COO early. The second is what the CEO needs to do. If the CEO's job is raising money and selling the vision, someone has to mind the store.
Seven types of COO
Across three editions, Nate and his co-author identified seven types. He walked through a few.
Heir apparent. The CEO is nearing the end of their tenure and the board wants to test drive a successor.
Two in a box. The CEO and COO are nearly inseparable. The classic case is Michael Dell and Mort Topfer in the early days of Dell. Dell knew the business was growing faster than his own development. He brought in Topfer, who was far more senior, and the two had offices with a glass wall between them so each always knew what the other was doing.
Change agent. Something heavy is coming, and someone needs to own it across silos.
The rest are in the book. For a CEO or board thinking about the hire, the first question is which type you're looking for.
Three kinds of fit
The second question is fit, and Nate separates it into three.
Executive fit is the chemistry between CEO and COO. Will they get to trust quickly?
Job fit is whether the COO's skills match what the job demands. A change agent and a mentor are different profiles.
Opportunity fit is whether the candidate's ambitions match what the role can offer. Many executives see the COO seat as a stop on the way to CEO. If the CEO is a founder who isn't going anywhere, that's a problem. When Mark Zuckerberg hired Sheryl Sandberg, he wasn't looking for a successor.
A note for HR leaders
Nate studies CEOs and COOs, but his point applies to every pair in the C-suite. As leadership teams grow, the number of those pairs grows with them, and the organization suffers whenever two of them aren't on the same page. If you're the one supporting a COO search, the conversation with the CEO starts with what they actually want from the seat: succession, skills, a partner, or something else.
How the role has changed
Each edition of the book marked a decade of pressure on operators. Between 2006 and 2016 it was the global recession and supply chain disruptions from natural disasters. Between 2016 and 2026 it was the pandemic, return to office, and now trade disputes and strained relationships between countries.
The data point that stands out: average COO tenure has dropped from around five years to about two and a half. Five made sense when the role was a proving ground for future CEOs. Two and a half is about half the tenure of the other officers in the C-suite, and Nate attributes it to how demanding the seat has become.
For a company going through growth or a turnaround, that's an expensive seat to keep refilling. It's worth getting the hire right, and worth doing what you can to extend the tenure once they're in.
Where to find Nate
Nate Bennett is at nate@gsu.edu and on LinkedIn. Riding Shotgun: The Role of the COO, third edition, is available wherever books are sold.
Enjoyed this episode? Please rate and subscribe to the Strategy and Leadership Podcast wherever you listen.
Work with Anthony
If your leadership team needs to get clear on direction, roles, and who owns what, Anthony facilitates strategic planning and alignment sessions with leadership teams across North America. He also speaks and runs working sessions at conferences and leadership events on strategy and alignment. Get in touch or book a conversation.
