Defending a Tradeoff

Senior signal is not picking the right answer. It is knowing what you gave up and being able to say so.

Every decision has a cost. Name it unprompted. "I'm choosing eventual consistency for the feed — the tradeoff is a user might briefly not see their own post, so I'll add read-your-writes for the author specifically." That single sentence demonstrates you understand the failure your choice introduces and have already handled it.

Distinguish pushback from disagreement. When an interviewer asks "why not X?", they are usually probing whether you considered X, not asserting X is better. Answer the probe: "X would work; I chose Y because our read/write ratio is 100:1 and X optimises writes. If writes dominated, I'd switch." You have now shown the boundary of your own decision, which is the actual thing being tested.

Change your mind for a reason, not for pressure. Caving the moment you are questioned reads as having had no reason to begin with. Refusing to update when given new information reads as worse. The distinguishing question is always: has anything changed about the constraints? If yes, update and say why. If no, hold and say why.

"It depends" is only acceptable with the dependency attached. "It depends on whether we need cross-region writes — if we do, multi-leader; if not, leader-follower is simpler." Naked "it depends" sounds like avoidance.

Admit unknowns cleanly. "I don't know the exact throughput of that engine; my estimate assumes low thousands of writes per second, and I'd verify before committing." Confidence about the wrong thing is far more damaging than a stated gap.