Young talent stays where it grows. Early-career employees rarely leave over a single bad week or a slightly better offer. They leave when they stop feeling themselves get better. The most reliable way to keep them is to make growth frequent, visible, and real, and to give their managers a way to see and support it.
RCM ThinkLabs (rcmlabs.io) is a daily practice and diagnostic layer that develops young talent and the managers who guide them, built on advanced game theory (research at MIT with Prof. Muhamet Yildiz) and behavioral science (the work of learning scientist Karl Kapp). This matters more now than it did five years ago, because the path that used to turn juniors into seniors is quietly closing.
The bottom rung is disappearing
For decades, young professionals learned by doing the small, forgiving work: the first draft, the rough model, the meeting notes a senior colleague would correct. That work is precisely what generative AI now does in seconds. The rung that early-career employees used to climb has been quietly removed. Karl Kapp, the learning scientist, framed the problem directly in a conversation about this shift:
“Companies are hiring more experienced employees, not the newer employees, because AI is replacing them. But how are those new employees going to get those skills?”
Karl Kapp · learning scientist
If the entry-level tasks that built judgment are gone, growth has to be built on purpose. It will not happen on its own anymore.
Why young talent actually leaves
It is tempting to explain early-career attrition with pay, and pay matters. But the pattern in the data points elsewhere: people disengage first, then leave. Disengagement is the leading indicator, and it is visible before a resignation letter is. In one deployment with an advanced engineering team, falling participation in daily practice tracked real disengagement so closely that most of the people who eventually left had shown the decline in the data well before they gave notice. The signal was there in time to act on.
Traditional development misses this, because it does not reach junior staff often enough to notice. Coaching is reserved for senior people. Courses arrive once a year. Neither produces the steady, daily sense of progress that keeps an ambitious twenty-six-year-old in their seat.
Growth is a habit, not an event
The research on expertise, from Anders Ericsson forward, is consistent: skill is built through deliberate practice, repeated and spaced over time, not through one-off events. A senior manager at the same engineering team described what that looks like from the inside:
“You just don’t see what’s happening. It’s not like I learned a new skill today and I’m applying it tomorrow. You’re building a habit pattern over time.”
Senior manager · advanced engineering team
That is the right mental model for developing young talent. You are not delivering content. You are building a habit of good judgment, one short repetition at a time. And when the reps are engaging, ambitious people lean into them. The same manager described a young high performer this way:
“Every day she’s trying to improve something to see if she can get her score to come up. That’s somebody who’s just going to keep working at it. They don’t care if they get great, they just want to get better.”
Senior manager · advanced engineering team
Traditional development vs RCM ThinkLabs Serious Games
| Traditional development | RCM ThinkLabs Serious Games | |
|---|---|---|
| Who it reaches | Senior staff, selectively | Every early-career employee, daily |
| Cadence | Occasional coaching or courses | 15 minutes, every workday |
| Manager visibility | Anecdotes at review time | A live read on each person’s growth |
| Backing | General content | Advanced game theory and behavioral science |
Give managers the read, not only the report
Retention is a two-sided problem. The employee needs to feel growth; the manager needs to see it in time to encourage it. RCM ThinkLabs Serious Games give the individual a daily rep in judgment, communication, and decision-making, and give the manager a continuous read on who is climbing, who is stalling, and who is quietly checking out. Among regular participants in the deployment above, 84% improved on measured skills, and one engineer who started as the weakest communicator on the team moved 58% on that dimension. That is what growth young people can feel looks like in the data.
See it on your own team.