Meta Senior Manager (M2) on Manager Career Growth, PIPs, Amazon vs Meta | Stefan Mai

Amazon's Culture and Early Management [0:00]

• Key Insights:

  • Amazon is described as having a ruthless performance culture.
  • Early managers might experience "death marches" requiring nights and weekends.
  • The advertising group at Amazon was a significant and profitable business unit.
  • Initial management roles can be challenging, with a steep learning curve and potential for mistakes.

  "So he's basically committing us to this death march for about you know six to eight months at the time and you know I had just left another job."

Early Management Mistakes and Developing Soft Skills [0:32]

• Key Insights:

  • Management is a "soft discipline" lacking clear output metrics, requiring strong storytelling skills.
  • A common early mistake is not spending enough time understanding individual team members' goals and objectives.
  • Building rapport and encouraging transparency from team members is crucial but can be difficult, especially when transitioning from a peer role.
  • Active listening, reflecting back what's heard, and asking curious questions are vital for understanding team members beyond a transactional level.

  "And ultimately, the same kind of information um is is is not all that useful. So, it's not enough to know that, you know, your friend is having a good time or a bad time. what you really want to know is deeper questions around you know what are their career objectives."

The Art of Storytelling in Leadership [11:51]

• Key Insights:

  • Storytelling is an essential skill for both managers and individual contributors to convey impact and gain buy-in.
  • Recognizing and critiquing one's own narratives is the first step to improving storytelling.
  • Managers should tell stories of team success to foster pride and motivation, and to build a foundation for future vision.
  • A winning team with a strong narrative often results in positive downstream effects on team dynamics and engagement.

  "So, you know, here's where we came from, here's where we are now, and here's how it's going to get us to the promised land. It's kind of the foundation of a great inspirational story."

Amazon's "Peculiar Ways" vs. Engineering Leadership [19:14]

• Key Insights:

  • Amazon's leadership model, historically referred to as "peculiar ways," emphasizes deep involvement at all management levels to track metrics and ensure discipline.
  • This approach can stifle creativity and innovation by requiring managers to be constantly aware of every detail beneath them.
  • This style is highly effective for growth-stage companies focused on incremental improvements (e.g., 1% revenue increase) but less so for exploration phases needing rapid innovation.
  • Kent Beck's "Explore, Expand, Extract" framework illustrates how different leadership styles are suited to different company stages.

  "But on the flip side, it can be really stifling to creativity and kind of like the ability to to work on on innovative new approaches if in some sense your bath needs to be queued in to all the things that you're doing."

Transitioning to Machine Learning and Career Growth [23:13]

• Key Insights:

  • Machine learning work is fundamentally different from traditional software engineering, involving shades of gray and requiring a scientific, experiment-driven approach.
  • Building intuition and practical experience through projects like Kaggle competitions is crucial for ML transitions, more so than just theoretical knowledge.
  • ML's ability to derive programs from data was a significant draw, offering a more inductive approach to problem-solving.
  • Complex ML applications, like those in Amazon's supply chain for demand forecasting and inventory prediction, showcase the intricate beauty of these systems.

  "But the the kind of like magic of it in some sense is being able to really execute. And I think that's missing for a lot of people."

Leading Other Managers and Organizational Dynamics [29:23]

• Key Insights:

  • Leading other managers is significantly more difficult than leading individual contributors, requiring different skills and awareness of complex power dynamics.
  • Managers can face "mutiny" where subordinates seek to displace them, often through subtle political maneuvering or by highlighting perceived incompetence.
  • In large organizations, especially those with senior ICs or multiple managers, gamesmanship and personal career motives can influence behavior, even if not overtly malicious.
  • Zero-sum mentalities often emerge in shrinking or stagnant organizations, contrasting with the abundance found in growing companies.

  "There is a very weird dynamic that unfolds amongst managers, leading managers. And I had a surface level grasp of this, but there was a lot more for for me to learn in order to be effective."

Comparing Amazon and Meta Cultures and Performance Systems [56:57]

• Key Insights:

  • Amazon is characterized by discipline and structured engineering design reviews, while Meta engineers are generally stronger but less disciplined, leading to more "fuzzy" reviews.
  • Meta is more open and collaborative, with engineers easily collaborating across global teams, whereas Amazon's approach is more process-oriented, relying on defined SLAs.
  • Meta's data infrastructure is more unified, providing easier access for ML work compared to Amazon's need to build individual data lakes.
  • Performance systems at both companies involve public ratings and private decisions on high-potential investments, but Meta is more aggressive in rewarding top performers with equity.

  "Amazon is disciplined and I think this is a a major strong point."

The Role of Low Performer Quotas and PIPs [01:05:36]

• Key Insights:

  • Companies of scale implement low performer quotas to counteract performance inflation and ensure high standards, often targeting 7-20% of employees.
  • Performance Improvement Plans (PIPs) serve as a legal paper trail but require a genuine chance of success to maintain credibility.
  • Managers often dislike implementing PIPs and firings, leading to difficult conversations and a desire for employees to succeed.
  • At Amazon, HR focused on checking boxes, while at Meta (historically), HR was highly protective of employee reputations, demanding exhaustive avenues before terminations.

  "And now you got to go change your tact. You got to go back to that person and say, 'No, actually, you know, you didn't kick ass that hard.'"

Career Growth, Job Hopping, and Mentorship [01:19:03]

• Key Insights:

  • Early career growth is often spurred by discomfort and new environments, making job hopping beneficial for learning and skill generalization.
  • At senior levels, frequent job changes can be a red flag, suggesting difficulty in achieving deep impact or building relationships.
  • Building credibility and influence through long-term tenure can be more advantageous than rapid movement at senior levels.
  • The most significant growth often occurs during transition periods, allowing individuals to battle-test skills and identify what is truly durable versus superficial.

  "I think if you look at the data for people who are like really ambitious in lower levels, moving around usually has a lot of advantages."

The Impact of AI on Interviews and Advice for Aspiring Leaders [01:12:24]

• Key Insights:

  • AI cheating tools are impacting interview prep, though candidates using them in mock interviews often don't benefit.
  • Companies are adapting by using AI-generated assignments or questions that AI struggles with, shifting focus to design and problem-solving over rote coding.
  • For companies like OpenAI and Anthropic, a deep understanding of functional areas and strong behavioral/retrospective interview performance are crucial, emphasizing technical achievements and walked-the-walk experiences.
  • Aspiring leaders should focus on identifying impactful problems, seeking mentorship, and actively planning their career trajectory, as personal agency is a key determinant of success.

  "And so, you know, the the kind of like noise and friction and everything else is going to dwarf any sort of gain that you might get."