Hiring Manager Behavioral Round

Standard HM BQ slot. 45–60 minutes covering 4–6 questions on conflict, mistakes, mentorship, and increasingly AI-usage philosophy. Same shape across SDE, MLE, EM tracks; EM gets deeper on people-management and decision-making, MLE gets a domain-knowledge thread woven in. Requirements 4–6 standard BQ-format questions; STAR-format answers expected. Recurring themes across recent loops: Conflict resolution with a teammate or stakeholder. A meaningful mistake / failure and what you learned. Mentorship / growing others. Influencing without authority / driving XFN alignment. Constructive feedback (giving and receiving). For senior candidates: how your performance was evaluated in past roles, and who set OKRs / how they were set. Why DoorDash (canned but expected). AI usage in your day-to-day — newer (2026) recurring question; multiple candidates report being penalized for shallow answers. Product-sense flavor for EM / SD candidates: "a customer complains about X — what could be the cause?" MLE variants can be almost entirely resume + values driven, with the interviewer drilling into methods used in the candidate's work rather than asking a fixed list of ML trivia. Notes Format: short 2-minute intro from interviewer, then 4–6 questions at ~7–8 minutes each, 5 minutes at end for candidate questions. The AI-usage question is the newest dimension. A weak answer is "I use ChatGPT for code review." A strong answer cites specific tools, specific workflows, measurable productivity changes, and the safety / quality safeguards you apply (e.g. always running tests, never trusting refactor suggestions without diff review). EM track: expect deeper drills on conflict (between a senior IC and a junior IC, between teams, between you and your manager), hiring (how you'd grow a team, how you screen), and difficult decisions (re-org, deprecating a product, downgrading a project). MLE track: expect a 10–15 minute domain-knowledge thread mid-round ("a user orders item X but receives item Y — what could cause this?" / "how would you prevent dashers from mis-delivering?"). Treat it as a structured-thinking question, not a buzzword quiz. Recent MLE loops may split the resume / values discussion and ML domain-knowledge discussion into separate rounds; do not assume the HM slot is only behavioral. Tone observation: multiple candidates report DoorDash HM interviewers being visibly disengaged or stoic. This is not necessarily a fail signal — interviewers are graded on consistency, not warmth. Pacing: several interviewers move fast through 6–7 questions and interrupt with "what was the result?" before you finish setting up the story. Keep each answer to 1–2 minutes and front-load the outcome — long setups get cut off. Senior loops increasingly drill project retrospectives with a product lens — if you redid this project, what would you change from a product perspective? Prepare a product-angle answer, not just an engineering one. Ownership-heavy short format Some Q3 2026 HM rounds ask only two or three primary questions and spend most of the slot drilling into each answer. The recurring axis is ownership: a problem you proactively identified, what you personally drove, the business impact, and how that impact was measured. AI usage remains a direct question. Keep the core stories deep enough for multiple follow-ups and reserve four or five thoughtful questions for the interviewer when the round leaves substantial candidate-Q&A time. Preparation Pre-write 6–8 STAR stories covering the recurring themes. Test them out loud; aim for ≤ 3 minutes each. Build one specific AI-usage story with concrete tooling, workflow, and outcome. Practice the safety / quality safeguard answer separately. For EM: build 2–3 hiring / firing / re-org stories with concrete numbers (team size, headcount change, attrition). For MLE: build 2–3 marketplace-flavored "why might X happen" answers; treat as structured reasoning, not memorization. Have 2–3 sharp candidate questions ready: team roadmap, on-call rotation, recent re-org impact, AI usage in the team.