Wider spans (15 or more direct reports per manager) encourage looser controls, greater independence, and more trial-and-error experiments. Which also leads to more failed experiments. Narrower spans (five or fewer per manager) allow tighter controls, more redundancy checks, and precise metrics. Which leads to fewer failures. There’s no right answer averaged across a company: we tailor the tools to the phase. When we assemble planes we want tight controls and narrow spans. When we invent futuristic technologies for those planes, we want more experiments and wider spans.