# CTO Leadership Through Experiments: Why Great Tech Leaders Don’t Make All the Decisions | CTO Teachings > How Fist-to-Five, One-Way Door Thinking, and Dictatorial Democracy Create Better CTOs Web page: https://ctoteachings.com/ai-engineering/cto-leadership-through-experiments-why-great-tech-leaders-don-t-make-all-the-decisions [All AI & Engineering articles](https://ctoteachings.com/ai-engineering) ## CTO Leadership Through Experiments: Why Great Tech Leaders Don’t Make All the Decisions Dean Hiller · July 16, 2025 This article is published in English only. 👋 *Welcome to CTO Teachings’s blog. While we are a recruiting company, we happen to provide a free blog for people just like you. It’s packed with hard-won lessons from CTO—who’s helped managers rise to Engineering Director and CTO roles.* **Let’s talk about what it really takes to become a great CTO. Or a great director. Or, heck, just a great manager.** One of the biggest shifts I see in **CTO leadership** or **engineering management** is learning to let go of control—and instead, build trust through structured experimentation. Let me walk you through something I use with all my teams. It’s called **Fist-to-Five**, and I pair it with Jeff Bezos’ **Type 1 vs. Type 2 decisions** (aka one-way doors vs. two-way doors). Together, they form the backbone of how I make decisions, run experiments, and empower teams. ### What is Fist-to-Five? It's simple. - 0 = Totally against it - 1 or 2 = Mild opposition, needs changes - 3 = Neutral-okay, with things to fix later - 4 or 5 = In favor Every decision starts with this. But there’s a catch—you need to qualify each decision as either a **one-way door** (hard to reverse) or a **two-way door** (easy to reverse). - **Firing someone?** One-way door. Tread carefully. - **Running a $500 marketing test?** Two-way door. Do it, learn, and adapt. ### CTOs Shouldn’t Make All the Decisions This is where **CTO coaching** really starts: breaking the belief that the leader must decide everything. I’ve had teams overrule me. I let them. Why? Two reasons: 1. I didn’t give them enough context. 2. I was just plain wrong. Either way, it creates a learning culture. But that doesn’t mean I never take control. Sometimes I’ll say, “We’re running this experiment. We’ll measure, then we’ll vote.” Like when I introduced **pair programming** every 6 weeks. People hated the idea. Three months later? They loved it. ### Creating a Culture of Rapid Experimentation This approach isn’t just about letting the team vote. It’s about weaving experimentation into every corner of how your team works: - Trying new ways of working together - Testing how you hire and grow the team - Balancing how time and resources are spent - Exploring different features or priorities - Asking: “What if we tried this for a while and saw what happens?” Every decision is a chance to try something, learn, and adjust. The trick is understanding the cost. What feels small and safe to test at one level of leadership might feel big and risky at another. You’ve got to get a feel for that—and adjust accordingly. ### When to Use Dictatorial Democracy Sometimes, you need speed. You don’t have time for long debates. That’s when I use what I call **dictatorial democracy**. I force a vote *before* people feel ready. Why? Because 80% of the time, they come in at all 4s and 5s. No need to waste 30 minutes discussing it. But the facilitator needs to be sharp. Especially with the 0s, 1s, and 2s. I’ve seen teams switch from all 5s to a full-stop 0 because one person pointed out something no one had considered. ### Real CTO Leadership Isn’t Loud I’ve had engineers say, “Dean, just make the call. Let’s move.” They don’t realize—if I do that, I lose the team. Sure, I gain one ally. But I disenroll 9 others. The better option? Convince the team. That’s **how to be a CTO** that earns long-term trust. The real flex is not being the hero—it’s building a team that leads with you. ### Closing Thoughts Want to grow into a **director of engineering** or **CTO**? Learn these three things: 1. Use **Fist-to-Five voting** for decisions. 2. Categorize decisions by **one-way vs. two-way doors**. 3. Embrace **experiment-led leadership**—even when it’s uncomfortable. Sometimes the team overrules you. That’s good. Either you didn’t give enough context—or you’re wrong. And in both cases, the team learns, and so do you. And yeah, sometimes you run the experiment first. You say, “We’re trying this one. Then we’ll vote.” You track how it goes. You bring the data back. And that’s how you get real buy-in. This works. Culture experiments. Hiring experiments. Budget experiments. Feature experiments. You name it. Just remember: when people say, “Just make the call,” they don’t see how it breaks the team. You might enroll one person—but disenroll nine others. That’s not leadership. The better option? Give them power. Let them decide. And when they’re not ready, vote anyway. Most of the time, they’re already at fours and fives. You just needed to ask. That’s how you build trust. That’s how you move fast. And that’s how you lead. *If you are ever looking for a job or expanding your team, connect with* *[Shelly](https://www.linkedin.com/in/shelly-h-761b6a358/)**,* *[Yuliia](https://www.linkedin.com/in/yuliia-hiller-952ab33bb/)**, or* *[Harold](https://www.linkedin.com/in/harold-alvear-83a552361/)* *on LinkedIn.* *Thanks for reading,* *CTO Teachings* [All AI & Engineering articles](https://ctoteachings.com/ai-engineering) [GET STARTED](https://ctoteachings.com/contact)