Your High Performers Have an Expiration Date. Most Companies Never Check It.

Three years ago, I assumed the employee with more time in their role was generally the one most ready for greater responsibility. I don’t make that assumption anymore. I ran on that assumption for most of my career. When I had to decide who got the harder project, the bigger team, the thing that could not fail, I looked at the internal résumé. Who has been here longest? Who has delivered before? Who knows where everything is buried? That instinct served me well for thirty years in enterprise software. Then the work changed faster than my assumptions about people did.

Here is the part that took me too long to say out loud. High performers now have an expiration date, and most companies never check it. The expiration date is not on the person. It is on the evidence we use to decide if they are still a high performer. People do not expire. Evidence does.
When you call someone a high performer, you are citing a measurement of how they did a job as that job existed at the time. That measurement was accurate. It may also be three years old, taken against a version of the work that no longer exists. Nobody re-dates it. It follows the person into every promotion conversation, and we keep making forward decisions on backward evidence.
That works fine in a stable environment, where past performance predicts future performance and tenure is a reasonable proxy for readiness, but AI broke the stability. The skills and experience that made someone indispensable yesterday could no longer tell me who would succeed tomorrow.

When we went AI-first at IgniteTech in 2023, I thought I was buying tools and training. What actually happened is that the work itself got rewritten. This is not a tech change. It is a culture change, and it changes what it means to be good at your job. A support engineer who used to escalate now resolves. A marketer who used to brief an agency now ships the asset. When work changes at that level, excellence at the old version of the job does not transfer automatically. Sometimes it does not transfer at all.

When Experience Becomes an Obstacle
This is the uncomfortable part. Time in the role may actually be more of a liability than an advantage, at least in the moments when the job is being redefined underneath somebody. The mechanism is simple. The more mastery someone has of an established process, the more they have to unlearn. Someone who spent eight years perfecting a workflow has eight years of good reasons why that workflow exists. They built it. They defended it in meetings. Their standing comes from being the person who knows how it works. Asking them to throw it out is asking them to give up the thing their authority rests on. Some do it and become the strongest people in the building. Others quietly protect the old system, and they are effective at it, because they have the credibility to make a patient case for waiting. Meanwhile, the newer person with no investment in the old process just starts working the new way because, to them, there is no old way. I watched that gap open inside my own company. People joining fresh became productive faster than veterans who first had to unlearn how they had worked for years. That was not about age or ability. It was about what they were carrying.

The Mistake I Made
The other half of this problem is not the employee. It is me. Leaders protect the people we are most afraid to lose, and that fear is almost always about institutional knowledge. They know the customer history. They know why the integration is built that way. So we wait. I waited. I told myself some of our strongest people would eventually come around, because they always had before. Some never did, and the waiting cost us more than the transition would have. Here is what I should have trusted instead. In thirty years I have never seen a company fail to recover from losing institutional knowledge. It hurts, and it carries real risk. It is also survivable, and it usually exposes something that should have been fixed years earlier, which is that the knowledge was never written down anywhere. We over-index on protecting that knowledge, even when doing so means waiting for someone to adapt while the work falls short of what the market now requires.

None of that means giving up on experienced people quickly. We gave people every opportunity to make the transition. For a full year before anything structural changed, we invested in bringing everyone with us. Up to $1,200 in education reimbursement for anything AI related, no pre-approval required. We paid licenses for the tools and dedicated twenty percent of the workweek to building with AI through AI Mondays. People came to me and said we only have four days now to do five days of work. I told them that was the problem worth solving. What I learned that year is that training videos don’t move people. Real projects with real deadlines move people. And the variable that predicted who made it through was not aptitude or seniority. We measured effort. We measured attitude. The difference was belief. If someone did not believe the work was changing, no tool or stipend reached them. Three years later, the results showed me the other side of tenure. In an anonymous company-wide survey this year, 89% of respondents reported moving up at least one level of AI fluency. When they joined, 72% rated themselves beginner or intermediate. At the time of the survey, only 4% did. And our longest-tenured employees reported the highest score on whether they would join again, at 100%.

That matters because this is not an argument against tenure. Tenure plus adaptation is the most valuable combination in any company. You get judgment, context and someone who can operate in the new environment. What no company can afford is unexamined tenure. Nine years in the same seat can also mean nine years without anyone asking whether the evidence we use to call that person a high performer still measures what matters. So I changed the questions I ask about my best people, and I would ask them about yours. Whose workflow actually changed? Not who talks about AI in meetings. Who has a measurably different Tuesday than a year ago? If you cannot describe how someone’s daily work changed, it did not. Who is producing something that was not possible before? The old question was whether we saved what we spent. The better question is, what can we do now that we could not do before? Some people on your team are already answering it with their output. Who shares and who hoards? Skills gaps close in months now. Hoarding is a choice.

Would you hire this person today into the job as it exists now? That is the hardest one. Ask it about your strongest people. Then ask it about yourself, because the CEO cannot delegate this and cannot be exempt from it. Run those questions twice a year. If you cannot answer them about someone, you already have your answer. The risk here is bigger than a bad call on one person. Companies survive bad calls. The risk is calcification. Old processes survive because senior people are invested in them. Promotions keep rewarding performance against work that no longer exists. From the inside, that looks like stability. From the outside, it is a competitor pulling away.
The real crisis is not mass displacement. It is mass irrelevance, companies that waited too long sitting next to competitors that did not. That does not start with technology. It starts with a CEO promoting off a measurement that expired. Check the date on your evidence. If it is older than the way you work now, it is expired.