When news broke recently that Google star developer Noam Shazeer had moved to OpenAI, I initially thought, “Well, high-value individuals are always going to jump ship.” But upon reflection, I realized there was something odd about it.
The fact that an individual exists whose departure could shake an entire company implies that there is already a list within that company of people who are “absolutely not to be lost.”
I became curious about who makes that list and how the company treats them once they are on it. I wondered if it’s a position that allows you to work more comfortably, or if it actually makes your work life even harder.
I happened to be reading an issue report from POSCO Research Institute (POSRI) based on interviews with Apple and Google officials, and I found an answer completely different from what I had imagined.
In domestic companies, “Key Talent” or “High-Potential” usually brings to mind receiving an employee award or going on an overseas training trip.
However, this report shows that Apple and Google’s approach is fundamentally different. Both have shifted the center of gravity for talent development from ’education’ to ‘actual work experience.’ Selection, development, and even disqualification all happen on the stage of real work.
How many people are selected as Key Talent?
Apple includes only 0.8% of its total workforce in this pool, while Google includes 1%. In terms of pure numbers, the probability is lower than being promoted to an executive position in a large corporation. That is how high the threshold is set.
Apple requires being in the top 10% of performance evaluations for three consecutive years as a base condition.
The scope covers all roles at the manager level and above at headquarters, excluding overseas regional organizations, and narrows down to core functions like hardware, software, AI, design, operations, and marketing. They manage a fixed scale of around 1,500 people out of approximately 186,000 employees—the 0.8% mark.
And one year of stellar performance isn’t enough. The point is that they require a three-year consecutive record. It means only those who have consistently excelled, rather than someone who had a lucky year, can even get near the list.
Google’s conditions are slightly different.
They require an ‘A’ grade in 3 out of the last 4 semi-annual evaluations over the past two years, along with experience in at least two different positions and two different business units (BUs). This means they value someone who has demonstrated performance across multiple environments more than someone who only did well in one department. Out of about 190,000 employees, roughly 2,000 are selected—again, around 1%.
What’s interesting is that this designation doesn’t start with a unilateral notification from the company. The door opens only when the individual expresses interest, followed by a process of aligning with the company’s expectations.
It’s not decided by one boss — Apple’s multi-stage collective validation
This is the part I found most interesting.
Apple does not decide this cutthroat 0.8% list based on the judgment of a single direct supervisor.
They conduct a verification process called the “Talent Planning Session,” held for two days per function, twelve times a year. The ‘function’ mentioned here refers not to a Business Unit (BU) but to professional functional organizations that cut across products, such as Hardware Engineering, Software Engineering, AI/Machine Learning, Design, Operations, and Marketing.
Apple doesn’t divide its organization by products like the iPhone or Mac teams; it operates on a structure where these functional organizations collaborate across multiple products. The key talent verification is also conducted thoroughly and regularly at this functional level.
On the first day, manager and senior manager-level peers who have worked together gather to verify performance and collaboration skills from a practical standpoint.
On the second day, higher-level leaders—senior managers and above—review the candidates to re-verify leadership and growth potential from an organizational and business perspective.
As the stages progress, the range of participants actually narrows. It is designed so that execution is judged by those who saw it up close, while strategic potential is judged by those who viewed it from afar—using different criteria. The candidates themselves do not participate in these discussions.
On top of this, they add a “Challenge–Defend” structure.
General participants take on the role of ‘Challenge,’ actively raising risks, limitations, and areas needing further verification for the candidates. The direct manager and senior manager who know the candidate well take on the role of ‘Defend,’ specifically justifying their performance and potential against those criticisms.
I wondered why they created such a cumbersome structure, and the answer was simple. The moment key talent is decided by the praise of only one person, the list stops being a list of top performers and turns into a list of people with good connections. Planting dissenting opinions at every stage of verification is essentially institutionalizing a habit of the organization questioning its own biases.
Why does Google require approval from leaders in other departments?
Google’s approach starts with the somewhat blatant premise that “our department’s bias will favor our own.”
Even if a candidate is recommended by their home BU, they go through a ‘Cross-validation’ where 5 Director-level leaders from unrelated BUs participate to re-verify leadership fit. It is a device to minimize bias, a phenomenon where subjective judgment from a specific boss or organization clouds the process. Requiring approval from leaders of other BUs means they treat “doing well in the current organization” and “doing well in a different organization” as separate questions from the start.
The criteria Google uses for this judgment are five standards commonly known as ‘Googleyness.’
Learning Agility: learning quickly in new situations,
Adaptability: responding flexibly to environmental changes,
Collaboration: exerting influence while working with other organizations,
Resilience: continuing to achieve results amidst failure,
Humility & Openness: accepting feedback and acting team-centrically.
None of these five items can be satisfied by the mere fact of “doing well right now.”
They are criteria that can only be judged by looking at how fast someone learns when moving teams, how they handle interacting with strangers, and their patterns for getting back up after failure. It means they view potential as the ability to reproduce performance in unfamiliar organizations or roles, rather than just doing the current job well.
They put performance and potential in different drawers to be judged separately, and I felt this distinction was quite cruel. Because even if you are doing well now, you are always subject to the suspicion that “you only did well because of this specific position.”
The real start is after being selected as Key Talent
If you’ve read this far, you might think, “So, doesn’t life get easier once you’re selected as key talent?” The answer is the exact opposite.
Apple throws high-difficulty strategic challenges that the company has been unable to solve for years at those designated as key talent. These are not educational case studies, but actual business problems, assigned not to individuals but to Cross-functional teams. They even exempt them from some routine work for about a year to focus on this task. It means they are observing the entire one-year execution process rather than short-term results. The task performance itself is a real-world verification arena that tests complex stakeholder management, cross-organizational collaboration, strategic judgment, and execution in actual business situations.
Google is no different. They assign tasks that are clearly more difficult than existing roles or Cross-functional projects that go beyond one’s area of expertise—actual ‘Live Business Issues’ rather than educational ones. The experience of clashing with new stakeholders and creating results in unfamiliar areas becomes the stage for testing Learning Agility and Adaptability.
Both companies have already shifted toward the direction that “work performance itself is development,” not “training followed by application.” In the past, the only reward for being selected as key talent might have been leadership training programs or external training, but now, the company places actual business headaches in that spot. Instead of giving a prize, they’ve added another test sheet.
5 pages of evaluation — How closely are key talents observed?
I noticed a figure while looking at Google-related materials.
For key talent, managers and skip-level managers write detailed evaluation reports spanning about 2.5 to 5 pages. Considering that general employee evaluations usually end in a paragraph or two, key talent becomes a subject of much deeper scrutiny, covering performance, leadership, potential, and development needs. This evaluation doesn’t just end with a grade; it is used as foundational data to design individual development programs and future roles.
For competency development, instead of external instructors, Senior Leaders recognized for their expertise within the organization act as instructors and coaches.
They structure 4–12 week programs focused on competencies used immediately in actual work, such as strategy formulation, persuasion and influence, and technical roadmap building. VPs and HR analyze individual evaluation results and weaknesses to recommend customized programs. It’s a method of prescribing different remedies for different people, which, in a way, also means the company is that acutely aware of each individual’s weaknesses. Some might call this meticulous care, but I felt ‘precise observation’ was a more accurate expression.
It stands out that it doesn’t end with a one-time lecture from a famous outside speaker, but rather a structure where people already verified within the company pass on their know-how through coaching. Since the instructor themselves has experienced the actual problems of the organization, the program content is highly likely to be methods that actually work within that company, rather than just theories. Consequently, key talent is observed through several pages of evaluation, while simultaneously receiving a personalized curriculum based on those very evaluations.
Key talent has two mentors and even management attached
At this point, it might seem like the pressure remains solely on the individual, but the company actually invests significant resources.
Apple assigns two ‘Guiding Mentors’ to each key talent. The main mentor is not the direct supervisor but a leader from another department, providing feedback unconstrained by the perspective of the home organization. Mentors, direct leaders, and skip-level leaders work together to create a development plan linked to organizational strategy. Furthermore, top management—such as the CEO, CFO, and Senior VP—personally check the progress of strategic tasks through a quarterly Leadership Committee. Excellent tasks are directly connected to actual business projects, and as one moves up toward the senior leader candidate pool, the intensity of management’s observation and verification increases accordingly.
Google uses Succession Planning itself as a development process.
Leaders at the VP level and above pre-designate multiple successor candidates for each lower level, managing multiple people rather than a single candidate. The candidate pool moves through steps: BU CEO candidate (Senior VP) → Alphabet Leadership candidate → Alphabet C-Level candidate. On top of this, they analyze the leadership trajectory by synthesizing multi-year ‘Stretch Assignment’ performance, supervisor evaluations, and external coaching feedback, including ‘Google Geist’ scores—an internal survey that regularly measures members’ perceptions of organizational culture and leadership. I found the fact that so many people were watching one person’s growth from so many angles at the same time a bit unfamiliar. It’s a density closer to surveillance than interest.
Up to 35% of key talent are dropped every year
And this was the most impressive part of the report: once you make the list, that position is not maintained automatically.
Apple conducts a ‘Pool in/out’ every year, and if expectations are not met, they are, in principle, removed from the list. However, they leave room for deferral for a certain period if there are reasonable grounds, such as temporary personal circumstances.
Google is even more explicit. They state that they maintain a pool turnover rate of 25–35% annually, meaning the list changes for about 1 in 4 candidates every year.
Apple also informs individuals about their selection as key talent privately to minimize friction within the organization, and while they offer differential rewards through additional salary increases and bonuses, they re-structure compensation conditions with a ‘Counter Offer’ to hold onto talent who express intent to leave.
Google similarly adds 8–15 percentage points to salary increase rates compared to general employees, and if they express intent to resign, they re-negotiate by adjusting conditions up to three times. They are cold-blooded enough to hold onto talent fiercely when they want to, but mercilessly remove them from the list if their performance falters.
The title of ‘Key Talent’ is not an ID card that stays with you once issued, but more like a certification that must pass renewal exams every time. This is exactly what both companies point out. They do not view falling out of the pool as a failure or a stigma, but rather leave the structure open so one can re-enter based on roles, performance, and growth stages. If you are used to an organizational culture where you rarely drop off a promotion list once you’re on it, this part will feel the most unfamiliar. I have laid out the structures of the two companies in the table below.
So, does being labeled a ‘Key Talent’ make life easier?
Returning to the question I first posed, the answer is clearly “No.”
From the moment you are selected, you are repeatedly verified by multiple leaders, saddled with problems the company hasn’t solved, scrutinized by management every quarter, and put back on the test bench every year.
This is not stable preferential treatment, but rather a position that comes as a set with constant observation and re-examination.
However, what you get in return is also clear.
Everything from compensation increase rates, promotion paths, touchpoints with management, to the difficulty and impact of the problems the company throws at you—everything accelerates in speed.
There was one thing that both Apple and Google confirmed.
The way to keep good talent is not to make things comfortable, but to keep giving them difficult problems and have many people watch the process together.
Whether this can be transferred directly to our company or our team is another matter.
Apple and Google have a workforce in the hundreds of thousands, so even if they pick only 0.8–1%, that results in a fairly large pool of 1,500–2,000 people, leaving room for multiple leaders to divide and verify them.
For smaller organizations, instead of copying this structure, it would be enough to adopt just two principles.
One is to not let promotions or selections be decided by the judgment of a single boss.
The other is not to give an easy position just because they were selected as key talent, but rather to assign them problems the company has not yet solved. That would be a much more honest verification than worrying about who to put on the list.
After reading this report, I had one thought.
It is better to let go of the expectation that the company will treat you comfortably the moment you hear the word ‘Key Talent.’
It is more realistic to see it as the moment the density of observation increases several times over. However, the story changes slightly if that observation is structured in a way where multiple leaders share the responsibility, rather than depending on one person’s whim or favoritism.
Whether you view that observation as surveillance or investment ultimately depends on this point.
And it is worth remembering that the qualification to receive that investment is not maintained from the moment you are selected, but must be proven again every year.
Ultimately, this position was closer to being on a permanent test bench. There was no moment where things became comfortable. That was the answer this report gave me.