Five product development steps Elon Musk applies
Build machines that build products
Early SpaceX Starship tanks were a maze of welds and joints, each demanding hours of painstaking work. Production was slow. In a review, Elon Musk posed the same question again and again: “Do we need this seam? This method?” Engineers examined every joint, uncovering many that existed out of habit, not necessity. They redrew the tank’s geometry, removed redundant welds and streamlined tooling. The design shifted from intricate to elegant: sections stacked cleanly, ready for rapid assembly. Build speed increased dramatically. Only then did automation make sense with welding rigs delivering precise, repeatable seams. By questioning, deleting, simplifying, accelerating and finally automating, SpaceX turned a slow, complex process into a fast, reliable production line.
Build using first principles thinking
If I had asked people what they wanted, they would have said faster horses. - Henry Ford
To build an exceptional product we need more than creativity. One that scales, survives and self-sustains needs the application of first principles thinking, driven by iteration and efficiency. Elon Musk’s five step approach to engineering and product development offers a pragmatic roadmap which shapes my approach.
1. Question every requirement
Make the requirements less dumb. - Elon Musk
Challenge every assumption about what’s necessary. If a requirement isn’t grounded in fundamental truths, e.g. laws of physics, treat it as suspect. Many startups die not from lack of innovation but from blind compliance to industry norms. Begin with a clean slate. Keep only that which survives scrutiny.
2. Delete before adding
The highest truth is to delete, not to add. - Robert Adams
If a part or step isn’t absolutely essential, delete it. The bias in most product development is to add (features, tools, options) but progress often comes from subtraction. What we remove defines the clarity of our offer and the elegance of the process. This is our trial-and-error phase. Cut ruthlessly and only add back what is undeniably valuable.
3. Simplify the process
Everything should be made as simple as possible, but not simpler. Simplify then simplify again. - Albert Einstein
Now we know what works, but it’s still rough. Time to simplify. A process that is too complex to understand is likely too complex to work well. Clarify every step, eliminate handoffs and redesign for flow. Use customer feedback as a key driver. A good product isn’t one that just functions, it does so predictably, consistently and with minimal friction.
4. Accelerate everything
I learned that the moment you want to slow down is the moment you should accelerate. - James Dyson
Once the process is clean and proven, shift from quality to speed. How quickly can we build, deliver and adapt? Time is a competitive advantage. Faster cycles mean quicker learning, higher customer satisfaction and greater throughput. Compress time where possible.
5. Automate intelligently
Taking an existing process and automating it does not make it better. It just makes a bad process faster. Fix the flaws first before automating. - McKinsey
Finally, automate only what’s already working. Never automate a process that hasn’t been simplified first. Doing so just hardwires inefficiency. Once our systems are lean and reliable, automate them using technology, e.g. software and robotics. The goal is to remove ourselves without losing control.
Other resources
How Elon Musk Solves Problems interview with Elon Musk
Elon Musk’s 6 Productivity Rules post by Phil Martin
Tips to Turn Startup Toys into Essential Tools post by Phil Martin
Elon Musk critically observers, “Possibly the most common error of a smart engineer is to optimise a thing that should not exist.”
Have fun.
Phil…
