Work Simulation and Work Style Surveys

After the coding section, Amazon's SDE online assessment has two multiple-choice parts: the Work Simulation and the Work Style surveys. 26 candidate reports from November 2025 to October 2026 mention them. This page summarises the format and what candidates say about their approach. It does not reproduce the assessment. Where we mention a pick, it is what one candidate chose, not an answer key. The posters say their own answers are not verified, and no one reports a score.

Timing

• The OA email gives typical times, which vary by report: 45 minutes for the Work Simulation and 15 for the Work Style surveys in two reports, 15 and 10 in an SDE II invitation. Candidates say neither part is strictly timed. One candidate did them two days after the coding.
• Some intern OAs have only the coding and a Work Style assessment.
• One SDE candidate (Django debugging version) also had a 5-minute personality test after the Work Style assessment.

Work Simulation: format

• It is set in a fictional team and looks like an email inbox. You read an email or a situation and either rate each possible response from 1 (least effective) to 5 (most effective), or choose what to include in your reply. One candidate describes ranking five options from best to worst. One poster warns not to flip the 1-5 scale.
• You play an engineer dealing with colleagues, a mentor and a team lead. One candidate played front-end, back-end and test roles responding to a colleague's bug report or code review.
• Some versions lean towards design. SDE II candidates describe design-related multiple choice, and decisions in low-level and high-level design scenarios. One recent report had one screen rating next actions on a project not yet started and another rating five ways to store votes.
• One SDE1 candidate (mid-2026) describes three modules, each a short project story with several scenarios. The first was about feedback and debugging on code you own; the second about a build-versus-buy decision and then starting a feature for launch.

Work Simulation: the kind of scenario (paraphrased)

1. A reviewer suggests a more complex design than yours because it pays off long term. What do you say back? The candidate rated offering to meet and hear the concerns highest, and telling the reviewer to change the code however they like lowest of the three.
2. Close to a review deadline you find an unexpected race-condition failure, and your lead asks for an update. Rate several first messages. The candidate rated flagging it now and asking for advice, with details to follow, highest; just announcing that the deadline will slip was rated lowest (1 or 2).
3. A teammate is still blocked calling your API before a morning status meeting. The candidate rated offering a set time to help today, and bringing in the lead if that fails, highest; pointing them back to documentation was rated lowest (1 or 2).
4. A teammate compares a vendor (faster launch) with building an internal platform (long-term cost and control). Rate how much each benefit should count in the recommendation.

The poster's own summary of the first module: accept feedback, troubleshoot with ownership, ask for help quickly, and do not dismiss a senior engineer's concerns. Of the second: understand what the launch needs, ask about technical requirements, protect long-term value, and raise timeline and code risks properly.

Work Style surveys: format

• Two surveys: one about how you approach software engineering work, one about your approach to work in general.
• One intern describes many personality-scale items plus forced-choice pairs: two statements side by side, and you choose on four points how much more like you one of them is (left Most like me, left More like me, right More like me, right Most like me).
• The pairs set traits against each other: planning, attention to detail, initiative, adaptability to change, learning and dependability. A typical pair (paraphrased): "I make a plan before I start" against "I deal with problems as they come up".
• One SDE1 candidate shared 35 such pairs with their picks, saying the survey seems to have no standard answer. When both statements were equally positive or equally negative, they tended to choose More like me.