Most leaders collapse performance reviews and talent assessments into one messy process. You sit down to evaluate someone’s year, trying to remember how they performed six to twelve months ago. As you spend an hour or more putting the review together, you also start to think about where they might be going next. Do they have what it takes to succeed in the next stage of the business? Are they ready for a promotion?
These are good questions, but trying to address an employee’s full-year performance and evaluate their potential at the same time is difficult to measure and ultimately communicate. The conversations get muddy. Your ratings become harder to defend in calibration meetings, and employees leave the review discussions unclear about where they actually stand.
Charles Thompson, an HR leader who redesigned his company’s entire review system, heard the same complaint from every employee he interviewed: the process took two hours per person and served no lasting purpose.
Performance reviews answer: How did this person perform this year?
Performance reviews cover the last 6-12 months. They focus on the person’s current role, what they were accountable for, how they delivered, and what outcomes they created. This conversation determines fair pay decisions, bonuses, and potentially promotions.
Example: If your goals were to close 50 deals and you closed 48, that’s performance. If you shipped a product feature by Q3 and you shipped it in Q2 with fewer bugs than expected, that’s performance. If you were accountable for reducing churn by 10% and churn dropped 12%, that’s performance.
Also important: If someone hears something in their performance review that they haven’t heard during the year, that’s a leadership problem, not a system problem.
Recency bias will destroy your data unless you collect feedback throughout the year
The last four to six weeks often overshadow the previous 11 months. One leader once admitted to me that he goes slow early in the year and ramps up in August and September because he knows that’s what matters for year-end conversations. That gaming of the system happens because leaders rely on memory instead of documented patterns.
That’s why many organizations are moving toward project-based feedback collection. Every time a project or deliverable is completed, data is collected from everyone involved. This creates an unbiased record of what happened six months ago, not just what you remember from the last month.
Talent assessments answer: What is this person’s trajectory and how critical are they for what’s ahead?
Talent assessments look one to three years out. They evaluate whether this person can grow in the ways the business will need, whether they elevate others, whether they’re shaping culture in a meaningful way, and whether they can stretch into bigger responsibilities as the company and its needs change.
Example: A startup was acquired by a large enterprise. Over the next two years, the HR system would change, the tech stack would change, and bureaucracy would increase. Leaders needed to know who on the team could thrive in a startup environment inside a larger enterprise. Some people had great skills but wouldn’t work well in that structure. Potential in this context meant adaptability and resilience through organizational upheaval.
Because high performance in the current role without the agility to pivot, learn, or stretch does not equal high potential.
Redefine your rating scale or it will work against you
Most organizations use a one to five rating scale, but the numbers mean different things to different leaders. A two might signal underperformance to one manager and “still developing” to another. Charles Thompson separated “developing” from “needs a performance plan” because those are two different situations requiring two different responses.
He also redefined the top ratings. A four became “exceptional,” meaning the person did way more than expected or did it way better. A five became “transformative,” which required answering a specific question: what did this person do that transformed the organization?
To be specific, if a leader can’t point to what someone did that transformed their team, department, or the company, they probably don’t deserve a five. Transformative impact might happen one year and not the next. You might work on a major project that launches in 2027, and that’s when your impact becomes transformative.
What makes this work is that if a leader rates someone a four or five, they must justify it with evidence and examples, not feelings or likability.
The nine-box model only works if you define “potential” clearly
The nine-box model plots performance on one axis and potential on the other. Someone can perform well but have low potential, meaning they’re good at the current state but may not grow with the organization. Someone else might have high potential but low performance, meaning they need support now but will be valuable later.
The problem is that five different leaders will define “potential” five different ways unless you make it explicit. Without calibration, leaders hoard talent by giving high performance ratings but low potential ratings so employees won’t get moved to other teams. Leaders also inflate ratings for people they personally like.
Example: One client defined potential based on who could handle the specific change ahead. They needed people who could thrive with increased bureaucracy, new systems, and constant pivots. Potential wasn’t about ambition or personality. It was about resilience, adaptability, and comfort with ambiguity during a two-year transition.
Takeaway: Before using a nine box model, spend significant time defining how you measure performance and how you measure potential, then require leaders to bring evidence and clear behavioral patterns, not just impressions.
Now What? Here are three immediate things you could do
Clarify your rating definitions. Write down what each number on your scale actually means and what behaviors or outcomes justify each rating. Share this with your team before reviews start.
Separate your performance list from your potential list. For each person on your team, write two separate assessments. One covers what they delivered this year. The other covers whether they can grow into what the business needs next.
Collect evidence now, not later. Open a document and write one concrete example for each person on your team of a decision they made, an outcome they created, or a moment that shows their impact. If you can’t think of examples, you’re not ready to rate them.
The work ahead
Performance reviews and talent assessments reveal how you think about your people, how you make decisions, and how seriously you take fairness. The models don’t do the work for you. You make the system work by being rigorous, collecting evidence throughout the year, and challenging your own biases.
