Chapter 25 · Free companion · A practice room
The round no other engineering interview prepares you for — and the one most candidates fail. It is also the most learnable, because almost nobody practises it and the thing being measured is a procedure rather than a talent.
An interviewer plays a customer. They describe a problem the way a real customer does — vaguely, in their own vocabulary, with the solution already assumed. There is no right answer. There is a right behaviour, and it is being scored.
You are not being graded on the plan you produce. You are being graded on how you got there.
Work it aloud, against a timer, ideally with someone playing the customer. Written answers do not build the skill — the round is verbal and the pressure is the point.
Before any solution, any architecture, any technology. Write only the questions you would ask. This is the drill that most changes interview outcomes, because solving too early is the most common failure and you cannot feel yourself doing it.
In this order. Under pressure people skip to move four; the skip is the failure.
Reconstructed from how these rounds are actually assessed. Score honestly after each practice run. Passing is roughly a 3 across the board with two 4s — but offers cluster where framing and evidence-seeking are 4–5, because those two predict field behaviour better than the rest combined.
| Dimension | 1 — weak · 3 — solid · 5 — strong |
|---|