In the early days of thinking about how talent contributed to organizational performance, the focus was largely on optimization. Frederick Taylor’s ‘scientific management’ aimed to measure and standardize every motion in search of the ‘one best way’ to do a task. When I talk about Taylorism, I often reference the scene in The Founder where the McDonald brothers choreograph the most efficient possible kitchen layout
.
Over time, this purely mechanical view was challenged. The Hawthorne studies showed that people responded not just to light levels, room temperature and break times, but to attention, meaning, and being asked what they thought. Peter Drucker’s “knowledge worker” and Deming’s systems thinking all of it pushed management away from control and toward stewardship.
Today’s research‑driven, tech‑enabled firms sit on top of that history. They run on knowledge, collaboration, and decisions made under uncertainty. One thing I see is that many firms still rely on fragmented systems: performance reviews over here, promotions over there, and learning as an optional extra. The result is predictable with lots of activity, very little coherence, and almost no ability to do real performance attribution for talent.
In a firm built on data and discovery, talent systems need to do three things well:
Connect performance, promotion, and development. People should experience one integrated journey with clear expectations, meaningful feedback, transparent promotion practices, and development opportunities that are obviously linked to their work.
Be credible to analytical minds. Frameworks must be grounded in evidence, informed by internal data, and co‑created with leaders. If quants and engineers wouldn’t trust a model built on vague assumptions, they shouldn’t be asked to trust a talent process built that way either.
Turn managers into stewards of growth. The real leverage point is the everyday interaction between managers and their teams. This impacts how they set goals, give feedback, nominate for promotions, and sponsor development. Systems either enable that stewardship or quietly undermine it.
It is not about owning a catalog of programs. It is about architecting the mechanisms by which performance is judged, promotions are decided, and learning becomes part of how the organization works.
In a place where the scientific method is part of the culture, talent development should look and feel familiar.
Hypotheses: If we change how we run performance reviews. This means having a critical view on what we measure, who’s in the room, how we calibrate. Further understand what happens to promotion outcomes and perceived fairness?
Experiments: If we redesign manager development for first‑line leaders in engineering and trading, can we see changes in team engagement, retention, or error rates?
Data and narrative: Not dashboards for their own sake, but metrics that help leaders tell a more accurate story about capability, potential, and risk.
The goal is not to turn people into data points. It is to use data to reveal where the system helps people succeed (and where it doesn’t) so that leaders can act.