Elon Musk uses a very specific order of operations when improving any system, product, or company.
Most people get this wrong by doing the steps out of order.
Here’s the correct sequence 👇
Step 1: Make the requirements less dumb
Every requirement is wrong some of the time especially if it came from a smart person.
- Question every requirement
- Ask who created it (not which department)
- Make the person responsible for defending it
- If no one owns it → it’s probably nonsense
“A requirement without a name attached to it is dangerous.”
Step 2: Delete parts or processes
This is where most companies fail.
- Actively remove things
- If you’re not adding things back ~10% of the time, you’re not deleting enough
- “Just in case” logic bloats systems endlessly
Bias should be toward removal, not addition.
Step 3: Simplify or optimize (ONLY after deleting)
This is the most common mistake engineers make.
- People optimize things that should not exist
- School trains us to answer bad questions instead of challenging them
- Never optimize before you delete
“You’re wearing a mental straightjacket if you optimize something unnecessary.”
Step 4: Accelerate cycle time
Only now should you speed things up.
- Yes, go faster
- But speed before simplification = chaos
- You can always make something faster later
Step 5: Automate (last, not first)
Automation comes after everything else.
Elon admits he’s personally failed by:
- Automating first
- Then accelerating
- Then simplifying
- Only to later realize the thing shouldn’t exist at all
True story:
Tesla deleted unnecessary components and bypassed a $2M robot simply by asking the right question.
The Core Insight
Order matters more than intelligence.
Most companies:
Automate → accelerate → optimize → keep everything
Elon’s order:
Question → delete → simplify → accelerate → automate
That’s how you build things that actually scale.
🎥 Watch the full video here: