Lately, I’ve been thinking quite a bit about where I want to take Decision Velocity next. Frankly, the way I’ve been talking about it has started to feel forced. The frame also feels narrower than the questions I actually want to explore. One thing I do know is that I don’t want to keep explaining authority, ownership, and decision closure as abstract ideas. They’re useful, but they come alive when there’s a real choice in front of us and enough evidence to think it through together. That’s what I was looking for when two Swiss company announcements recently caught my attention. Why Swiss companies? No particular reason. They just happen to be. On Holding moved toward co-CEOs. SoftwareOne moved away from them. That contrast was enough to make me look more closely. At first, it sounds like another invitation to debate whether two CEOs are better than one. I’m not very interested in that debate. What kept me looking was the reason each company gave for choosing its structure. The closer I looked, the more the two announcements seemed to belong together. So why these two companies? If you don’t know On, it’s the Swiss performance sportswear company behind the Cloud running shoes. It was founded in 2010 and has grown into a global footwear, apparel, and accessories business. Its 2025 sales surpassed 3 billion Swiss francs (CHF), and it now operates in more than 90 countries. SoftwareOne, on the other hand, is a very different business. It helps organizations buy, manage, optimize, and modernize their software and cloud technology. In July 2025, it combined with the Norwegian company Crayon. The enlarged company had about 13,000 people across more than 70 countries. So no, I’m not putting them together because running shoes and cloud software have much in common. I’m putting them together because both are Swiss-headquartered global companies dealing with a new level of complexity. Both recently used co-CEOs. Both talked about leadership structure as an answer to the phase the company was entering. And within less than four months, one moved toward shared executive leadership while the other moved away from it. That’s a useful comparison. We get to see the same visible tool used for two very different jobs. What caught me at On In March 2026, On announced that co-founders David Allemann and Caspar Coppetti would become co-CEOs. They’d also remain Executive Co-Chairmen of the board. Scott Maguire would become President and COO, with responsibility across the full value chain, from research and manufacturing through marketing, commercial operations, and technology. The titles are interesting, but the wording behind them is what caught my attention. On said it wanted to connect “founder-led strategic intent with execution,” preserve entrepreneurial speed, and align founder stewardship directly with the work of running the company. One year earlier, On had moved from co-CEOs to Martin Hoffmann as sole CEO. Now it was changing again, this time bringing two founders into the co-CEO roles and placing a President and COO across the operating core. You could look at that sequence and say On hasn’t decided what structure it wants. I don’t think the public record gives us enough to say that. What I see instead is a company moving judgment around as its needs change. The 2026 design seems to bring founder judgment closer to the top operating role while giving one executive a broad view across the value chain. That’s an important tension in many founder-led companies. How do you keep the judgment that made the company distinctive without making every important operating decision travel back through the founders? The announcement tells us what On wants the structure to accomplish. It doesn’t tell us how the authority works when the choices get difficult. Imagine a product launch where brand ambition, manufacturing readiness, margin, and a market commitment all collide. Which part belongs to the co-CEOs? Which part belongs to the President and COO? When does asking for input turn into asking for approval? If the three roles see the trade-off differently, who can close the call? I don’t know how On has answered those questions internally. It may have answered them very well. My point is that those answers, not the titles, will determine whether the structure gives the company the agility it says it wants. SoftwareOne made the opposite move SoftwareOne had a different problem to solve. When it combined with Crayon in July 2025, Raphael Erb and Melissa Mulholland became co-CEOs of the enlarged company. Erb looked after commercial operations, services, and the marketplace. Mulholland handled strategy development, customer platforms, and global functions, among other areas. You can see why that might be useful during an integration. Two global businesses were bringing together leadership teams, operating models, customer relationships, brands, systems, legal entities, and regional organizations. Shared leadership could keep knowledge and continuity from both sides in the room while the new company was being assembled. Then, in July 2026, SoftwareOne said the integration was substantially complete and named Erb sole CEO for the next phase. This is the part I keep coming back to. SoftwareOne’s co-CEO structure had a stated job. It was there to support continuity through an integration. When the company believed that job was nearing completion, it changed the structure. In other words, the design had something close to an expiry condition. I can’t tell you from the outside how well the co-CEO model worked. I can’t tell you whether moving to one CEO will improve execution. I also can’t see how the knowledge and relationships held across two roles will transfer into one. What I can say is that the public explanation makes the logic visible. The structure wasn’t presented as a permanent belief about leadership. It was an answer to a particular phase. I think more companies would benefit from being that explicit. Here’s the idea I can’t shake I’ve started to think of a leadership structure as a hypothesis about where judgment needs to sit right now. On’s apparent hypothesis is that founder judgment and an integrated operating core need to sit close together as the company scales. SoftwareOne’s integration hypothesis was that shared executive leadership would help two global organizations become one. Its next hypothesis is that a sole CEO will better serve focus and execution in the phase that follows. Neither choice tells us that centralized authority is better than distributed authority. That’s too simple. I’ve been using another phrase in my own thinking: the smallest complete authority. I don’t mean the lowest person on the org chart or the fewest people in the room. I mean the smallest person or group that holds enough authority to make the whole trade-off and let the decision move. Sometimes that’s one person. Sometimes it really does require two people with different judgment. Sometimes it belongs to an operator working inside a clear boundary. The trouble starts when responsibility sits in one place but enough authority to make the trade-off sits somewhere else. Then the decision has to travel again. It gets translated, softened, checked, reopened, or quietly routed back upward. That’s where the org chart stops being a picture and becomes a decision system. Four questions I’d want to ask If you and I were inside either company, I wouldn’t start by asking whether the structure looked modern or unconventional. I’d want to talk through four things. What job is the structure here to do? Not just “why do we have two CEOs?” What important kind of decision becomes easier to make because leadership has been designed this way? At On, the public answer seems to involve connecting founder intent, product, brand, and global execution. At SoftwareOne, the co-CEO answer was integration and continuity. The sole-CEO answer is focus and execution in the next phase. If you can’t say the job plainly, the structure may be carrying history rather than a current operating purpose. Which decisions are actually shared? “Shared leadership” doesn’t mean every decision should be shared. There’ll be cases where two forms of judgment genuinely belong together. There’ll be others where one person needs to make the final call. If nobody makes that distinction, ordinary work can turn into a negotiation among titles. Who owns the collision? This is where the chart meets the real work. A product ambition runs into manufacturing capacity. A customer-platform choice changes a commercial commitment. Each leader can be perfectly clear about their own area while nobody owns the trade-off between them. Sometimes the most important authority rule isn’t who owns each box. It’s who owns the collision between them. What would make us change the structure again? SoftwareOne connected its co-CEO design to an integration mandate. That made the later change understandable. Other structures stay in place long after the reason for them has faded. People keep routing decisions through a design built for an earlier phase, then make up for the gaps through private conversations, repeated approval, and informal workarounds. You don’t need an expiry date stamped on the org chart. But you do need some idea of what would trigger a review. It might be the end of an integration, a change in scale, a new strategy, a different role for the founder, or the same decisions repeatedly getting stuck between the same people. Why this sits inside Decision Velocity Decision Velocity started with a fairly simple observation … a decision can sound settled in a meeting and still fail to become usable in the work. The same thing can happen with organizational structure. Giving someone a CEO, COO, or President title doesn’t finish the decisions about authority underneath the role. People still have to k
