For engineers: how to evaluate an offer without over-optimising for comp

We debrief roughly two hundred senior engineering hires a year. The regret pattern is remarkably consistent: at year one, engineers who changed jobs are pleased about the compensation. At year two, they wish they had asked more questions about scope, manager, and codebase health, in that order.

The four variables that actually matter

• Scope — the specific system or product surface you will own. A vague answer here is the single strongest predictor of regret.
• Manager — the person you will report to for the next 18 months. Their reputation, their retention record, their willingness to be candid in the interview.
• Codebase — the health of the systems you will inherit. Ask to see a real PR review, not a marketing tour.
• Compensation — the number, the structure, the vesting cliff, the refresh cadence. It matters, and it matters less than the first three.

The questions to ask before you accept

The best diagnostic question we have found for evaluating a new manager: 'Tell me about someone on your team you promoted in the last twelve months. What did you do specifically to make that promotion possible?' A manager who cannot answer this in specifics has not been paying attention to their team; a manager who can is showing you their operating model.

The bar for leaving

Do not leave for a 20% raise. The switching cost — the six-month ramp, the lost equity vest, the political capital you had built — is roughly equal to a 20% raise in the year one economics, and you take on new risk. Leave for a step-change in scope, for a manager you have wanted to work for, or for a company whose next chapter you actively want to be part of. If none of those are true, negotiate where you are.

"The best career moves compound over five years. The best compensation moves compound over eighteen months. Choose accordingly."