Posted by
Ask an engineering leader why they can’t just use a standardized evaluation for their hiring, and you will hear some version of the same answer. Our context is different. Our stack is specific. What works here wouldn’t work elsewhere.
Every company believes this. Which should be the first clue that something is off.
I have spent ten years looking at how companies define a good engineer. Startups, GCCs, product companies, services organizations. The definitions overlap far more than anyone inside those companies would admit. Strip away the vocabulary each company uses, and what remains is remarkably consistent.
- Can this person reason through a problem they haven’t seen before?
- Can they write code another human can read and maintain?
- Can they make sensible tradeoffs under constraints?
- Can they communicate what they are doing and why?
- Can they learn?
That is most of the bar. At most companies. For most roles.
The genuinely company-specific part exists, but it is much smaller than companies believe. Familiarity with a particular domain. Comfort with a certain pace or ambiguity level. A specific depth in one part of the stack. Real, but narrow. A layer on top of the foundation, not the foundation itself.
Yet companies behave as if the entire definition is proprietary. Every organization builds its own interview process from scratch. Writes its own questions. Trains nobody. Calibrates against nothing outside its own walls. Thousands of companies independently reinventing the evaluation of the same underlying abilities. Each convinced their version measures something the others don’t.
This also explains why hiring resisted being productized for twenty years. You cannot build infrastructure for something every buyer insists cannot be standardized. So the industry built around the belief instead of questioning it. Sourcing tools, scheduling tools, tracking tools. Everything except the evaluation itself.
We accidentally built the proof otherwise. Evaluating engineers across every segment for a decade forces you to see the pattern. When the same candidate signal predicts success at a fintech startup, a retail GCC, and an enterprise product company, the story that every bar is unique becomes hard to sustain. The foundation transfers. It always did.
None of this means companies should stop caring about fit. The specific layer is real and worth assessing carefully. But the industry has the proportions backwards. Companies spend most of their process re-testing the universal foundation. Badly, with untrained interviewers and uncalibrated questions. Then they rush the genuinely specific part into a thirty minute conversation at the end.
Get the proportions right and hiring changes shape. The universal foundation can be evaluated consistently, once, in a way that transfers. The company-specific layer gets the focused human attention it deserves.
The bar was never as unique as we told ourselves. The sooner the industry admits that, the sooner hiring gets its infrastructure.