Bottlenecks (Systems)

You're seeing it when your engineering team ships features fast but every release waits two weeks for a single QA engineer's approval. The bottleneck isn't engineering velocity — it's the QA stage that governs total throughput.

You're seeing it when you double your marketing spend but revenue barely moves because your sales team can't handle more inbound leads. The bottleneck is sales capacity, not demand generation.

Identify the single biggest constraint in your company right now. Is it hiring? Sales capacity? Product velocity? Ship speed? Focus disproportionate resources on that one constraint. Everything else is noise until the bottleneck moves.

Musk's approach to Tesla manufacturing was a masterclass in bottleneck thinking. During Model 3 production hell, he personally moved onto the factory floor to identify constraints. He discovered the bottleneck wasn't the main assembly line but the battery module production in the "tent" — a single sub-process that governed the entire factory's output. He redirected engineers, redesigned the process, and slept at the factory until throughput at that stage improved. Once the battery bottleneck broke, overall production surged. Founders should learn: when output stalls, resist the instinct to push everything harder. Find the one constraint, break it, then find the next one. Systems improve sequentially through their bottlenecks, never uniformly.