Ethan Evans on Extending the Magic Loop, Systematic Inventiveness, and the Amazon Failure Story

Source:
Ethan Evans on Extending the Magic Loop, Systematic Inventiveness, and the Amazon Failure Story
transcriptlenny-podcastcareerleadershipamazoninvention

Ethan Evans on Extending the Magic Loop, Systematic Inventiveness, and the Amazon Failure Story

Ethan Evans, former Vice President at Amazon and executive coach, in his second conversation with Lenny Rachitsky. The episode revisits The Magic Loop (now the sixth most popular post in Lenny’s newsletter history), extends it to executive level, covers systematic inventiveness, and tells the story of publicly failing Jeff Bezos on the Amazon Appstore launch.

Key ideas

  • At executive level, the Magic Loop becomes invisible. You stop asking “how can I help?” and start anticipating. Directors and VPs proactively address problems and keep leadership informed. The loop runs continuously; you just no longer announce each cycle. Premature escalation to this mode without established trust backfires.
  • Invention is two hours a month. Systematic inventiveness does not require genius — it requires dedicated, distraction-free thinking plus a method: combine two things from different domains. Once you have a good idea, it takes years to express it. You do not need many ideas to be seen as highly inventive.
  • Own the failure loudly and immediately. When Jeff Bezos’s app store launch failed, Ethan said “it’s my fault” and bought time in one-hour increments with status updates. The lesson from Jeff Wilke: if you know you are gambling and lose, you can recover. If you did not know you were gambling, you probably should not still be in the role.
  • Meeting in person resets flame wars. Moving an email argument to a face-to-face meeting changes the dynamic: it is hard to be visibly angry at someone sitting next to you. Two years after the failure, Ethan was promoted to VP.
  • “What your manager should do and $4 will get you a cup of coffee at Starbucks.” The Magic Loop puts career control in your hands rather than waiting for a good manager to notice you.

The Magic Loop revisited

The five steps: (1) do your current job well, (2) ask your manager how you can help, (3) do whatever they ask, (4) ask for something that advances your own goal, (5) repeat.

Two additions from the second episode:

The manager-initiated form. Leaders can start the loop from their side: “I will invest in your development if you step up to new challenges. What are your goals?” Ethan found this created strong alignment — employees lean in harder when they see reciprocal investment is real.

Evolution at executive level. At VP level, you stop asking. You anticipate, act, and inform. The explicit conversation shrinks to nothing; the underlying pattern does not. Getting to that mode requires the trust built through years of cycling through the earlier steps.

The senior manager bottleneck

Senior managers get stuck because: (1) there are simply fewer director roles than senior manager roles — a natural choke point; (2) the skills that got you to senior manager (individual execution, functional strength, getting things done) are not the skills required at director and above (influence, coordination, strategic thinking, letting go of details).

What Got You Here Won’t Get You There (Marshall Goldsmith) is Ethan’s recommended read for this transition.

The practical advice: start practising next-level skills now. Take on strategic projects. Show you can operate without being in all the details. Make yourself the person who gets given teams when the organisation is rationalised, not one of those whose teams get redistributed.

Systematic inventiveness

Ethan holds 70+ patents accumulated over 15 years at Amazon. The method:

  1. Become an expert. You cannot invent without a knowledge base. Fumbling in an unfamiliar domain produces nothing.
  2. Block dedicated time. Two hours per month, no devices, focused concentration. “People feel like invention is just going to come to them.” It does not.
  3. Combine two things from different domains. Ethan’s drone-delivery patent: he loved military history, thought about aircraft carriers, applied the concept to trucks. A truck with no roof drives slowly through neighbourhoods; drones fly back and forth from the truck to houses. The idea arose by combining drone delivery (his domain) with aircraft carrier logistics (adjacent domain).

One good idea takes years to express. Prime is a 20-year-old idea, still developing. The Kindle is decades old, still improving. You do not need many ideas to be seen as inventive; you need to pursue them rigorously.

The Bezos app store failure

The Amazon Appstore launched with a feature Jeff Bezos had personally championed (Test Drive — simulate any app in a browser before downloading). The feature was not working at launch. At 6:15 AM, Jeff asked where the letter was. Ethan admitted the problem. A snowballing email thread followed, with Jeff’s number two and skip-levels copying in.

What Ethan did:

  1. Owned it immediately: “Yes, it’s not working. It’s my fault. I will deal with it.”
  2. Gave hourly status updates: “Here is where we are. Here is the plan. Next update at [time].”
  3. Accepted help: Andy Jassy’s AWS team sent principal engineers who threw 500 machines at the crappy database design.
  4. Met Jeff in person the following week — sat next to him voluntarily, expecting a reckoning, and instead got “How are you doing? I bet it’s been a hard week.”

Two years later: promoted to Vice President.

Jeff Wilke’s verdict: “At least you knew you were gambling. If you hadn’t known you were gambling, we’d be discussing your departure.”

Amazon leadership principles relevant to this episode

  • Ownership: “An owner never says that’s not my job.” Ethan contributed this phrasing to Amazon’s official leadership principles.
  • Bias for Action: Commonly misread as “make big bets.” Correctly: do small, reversible things to learn. The bias is toward action over analysis paralysis, not toward recklessness.