Becoming a star engineer—Part III

Research that identifies a set of success behaviors is only half useful. The other half is the question nobody wants to ask too loudly: can these behaviors actually be taught, or do they just describe people who already had the underlying disposition?

Kelley’s team spent seven years answering that question. The study tracked over a thousand engineers across US and European companies through a structured training program—what they called the Breakthrough program—designed to teach the nine work strategies described in Part II. Managers rated participants on productivity factors at the start and again eight months after completion.

The results were unambiguous. Engineers who went through the program solved problems faster, produced higher-quality work, and earned stronger evaluations from customers. The strategies weren’t just descriptive of stars—they were causally connected to star performance, at least when learned and applied consistently.

One finding deserves particular attention. About 30% of participants were already rated as high performers before the program began. The training worked for them too, but the most dramatic gains came elsewhere. Women and minority engineers—groups that often face additional structural friction in technical organizations—showed average productivity improvements of 400%. That number is large enough to require some interpretation.

Part of what the program likely provided to these participants wasn’t just strategies but permission. If you’ve spent years in an environment that subtly signals your judgment is less trustworthy, your ideas less worth backing, your political capital thinner than a peer’s—you adapt. You self-censor. You under-invest in the relationship-building and initiative-taking that stars do instinctively, because experience has taught you those investments don’t pay at the same rate. The program created a context where those behaviors were explicitly valued and practiced. The productivity gains followed.

The broader finding, though, applies across the full participant pool: when engineers underperform, it’s rarely a capability problem. They’re typically smart enough, technically prepared enough, and motivated enough. What they’re missing are the work strategies. Not because they couldn’t learn them—because nobody ever taught them. Engineering education covers algorithms and systems and languages. It covers almost none of what the research identifies as the actual determinants of career-level performance.

That gap is a choice, and it’s a correctable one. The question worth sitting with is whether your organization is making it deliberately or just by default.