Skip to content

Myth of a superstar employee

We've all heard the business adage - "A players hire A players, B players hire C players." Our generation was taught this concept by Steve Jobs [1]. However, organizations often take this too far, anointing a chosen few as "high potential" stars that receive the best projects, promotions and perks. This perceived favoritism can be toxic.

graph TB A["Leadership"] --> D{Assign Quality <br> Work Equally} B["High Performer"] --> C{Support Others' <br> Success} D --> |No| H([Team Eternally in Crisis Mode]) D --> |Yes| G["Lower Performers"] C --> |Yes| G C --> |No| F([Lower Engagement]) F --> |Leads to| E([High ego and/or burnout]) G --> |Boosts| K([Team Morale & Productivity]) G --> |Improves Skills of| G


What most leaders of large companies miss today is:

  • Steve Jobs made that comment about a time when computer science was incredibly hard. You had no giants to stand on. Today you have many giants who have produced limitless content on the world-wide web. Additionally, we now have high-level languages, libraries, APIs, standards, university degrees, and an entire world of support to find and implement best practices.
  • Steve Jobs often also talked about teamwork [2]. He frequently mentioned that ideas get polished when people bounce them off each other and collaborate over a long period of time. He talked about building a great team, and not just finding A players.

I think most companies forget that second step. It is also only through that second step that you can nurture your team members to function as A players for any sustained period of time.


Research shows that when average employees believe a privileged inner circle of "superstars" exists, engagement and retention fall across the entire workforce [3]. Resentment spreads like a virus, as most conclude, "Why bother going the extra mile?" Employees learn to refer to their employer and their team as "them" instead of "we". This I say from experience – for example, I once asked a friend what his team is building and if he likes his new office that his team was moved into. His answer was entirely in the 3rd person with "they" and "them". Many have anecdotally attributed high attrition to this problem of star culture and lack of engagement.


In my experience, we leaders should expect rockstar contributors to coach and develop talented newcomers. This gives these higher performers new purpose while improving camaraderie and trust company-wide. Studies also find that star employees themselves benefit from collegiality despite their dislike for it; working among supportive peer groups guards against ethical blindspots that high-pressure environments can otherwise enable. Pride comes before the fall.

The sustainable path lies not through superstars but "superteams" - groups that maximize flexibility, development, and knowledge sharing for all contributors, not just those handpicked for the fast track.

Here are four building blocks for more balanced talent management:

  1. Distribute High Visibility Projects - Make sure all competent team members, not just the showiest, get a chance to lead or address stakeholders.
  2. Invest In Strugglers - Develop adequate performers rather than rushing to ax "underachievers." Patience and support usually helps retain institutional knowledge and save a lot of costs.
  3. Embed High Talent - Have strongest collaborators train others via mentoring and informal sharing. Avoid siloing them among fellow stars.
  4. Make helping others a currency - Ensure that you don’t ask high-performers to thanklessly contribute towards their team’s success. Everyone’s success should be everyone’s business, and in order to make this happen, the performance evaluation system should have a strong weightage towards influence and communication.

Think rock-climbing teams, not 4x100m relay race teams.


I would like to give you something actionable that has helped me implement these ideas and achieve significantly high team retention for my employers. I have developed a standardized performance scoring system that can help you to create these superteams:

Category Skill Weight Description
Technical Skills Coding/Engineering Expertise 20% Proficiency in programming languages, databases, and cloud platforms relevant to data science and machine learning.
Mathematical Expertise 20% Ability to understand and apply statistical and mathematical concepts to data analysis and modeling.
Task Lengths/Code Simplicity 10% Ability to break down complex tasks into manageable steps and write clean, maintainable code.
Soft Skills Influence 20% Ability to effectively communicate and collaborate with technical and non-technical stakeholders.
Pace 10% Ability to work efficiently and deliver results on time.
Work Review Preparedness 10% Ability to come prepared to code review and math logic review sessions and meetings.
Communication 10% Ability to clearly and concisely present technical information to both technical and non-technical audiences.
  • Focusing on objective criteria: The system focuses on specific skills and behaviors that are important for success in a data science and machine learning organization, rather than relying on subjective judgments or gut feelings. On each criteria, the categorical scores of 1, 2, and 3 are simply sink, flow, and source, respectively.

    • Sink: They need help in this criteria
    • Flow: They are able to draw from available resources for this criteria
    • Source: They are an example to the team on how to ace this criteria
  • Providing a framework for self-improvement: The system can be used to identify areas where individual team members would like to improve, and to create development plans that match their career goals. There is room here for a team of diverse skill sets. Afterall, you should never judge a defender by the number of goals they have scored for their team.

  • Promoting collaboration: The system emphasizes the importance of communication, collaboration, and influence, which are essential for building strong "superteams."
  • Encouraging knowledge sharing: The system rewards team members who are willing to share their knowledge and expertise with others, which can help to break down silos and create a more cohesive team environment.

Finally, once a team member has been scored on each criteria, a weighted average can be calculated based on any weights you feel fair for your team (mine are shown in the table above). I also judge leveling within my team using these scores. I use the following average score guidelines for leveling:

Level Average Score
Junior/Entry-level 1.5
Mid-level 2
Senior 2.5


The myth of the indispensable individual superstar still pervades too many HR practices and team cultures, disengaging the broader workforce. However the model I have presented in this article should empower leaders to in turn empower their team members, and create an environment for building great products that team members enjoy and want to sustain.


[1] Steve Jobs' comment about A-players. Apple Wiki, Fandom, 1995. link

[2] Williamson, Mark. Steve Jobs: Teamwork | Greatest Leaders, 2010. link

[3] Sull, Donald, et al. “Why Every Leader Needs to Worry About Toxic Culture.” MIT Sloan Management Review, 2022. link

While writing this article, I stumbled upon a book, Superteams by Khoi Tu, that seems to share my concept of 'superteams', if you go by the synopsis and the sample pages available on B&B Nook.

Tu, Khoi. Superteams: The secrets of stellar performance from seven legendary teams. Penguin UK, 2012. (B&B Nook, Google Books)

Some relevant quotes:

  • "The myth of the single hero is just that: a myth"
  • "Many people think of it like a rock supergroup: bring the best of the best together and magic will happen. Yet supergroups often flop, while bands of unknowns rise to the top."

This book will also add to my reading list, as I have so far only read the context and not any of the 7 stories detailed within.