What is the goal of contextual inquiry and what do engineers gain?

WHAT THIS TESTS: Whether you understand that early discovery must be grounded in observed behavior rather than self-reported summaries. Senior engineers are expected to know that users omit low-level details, workarounds, and environmental constraints when they are removed from the actual task. The question probes whether you see ethnographic field research as a legitimate input to system architecture, integration decisions, and reliability engineering. It also checks if you recognize the difference between recall-based methods like surveys and contextual methods that capture invisible habitual actions.

A GOOD ANSWER COVERS: First, the primary goal is to uncover hidden work practices and tacit knowledge by observing and interviewing users in their natural context while they perform real tasks. Second, you should expect to see environmental realities such as interruptions, multiple monitors, paper notes, and cross-tool workflows that never appear in lab studies. Third, you should look for workarounds and superstitious behaviors like habitual saving despite autosave or manual data reconciliation between systems; these signal trust gaps and integration needs. Fourth, you should connect these observations back to engineering decisions such as data models, API boundaries, offline handling, error recovery, and performance requirements. Mentioning that you would attend to hear verbatim comments and see sequence timing is also strong.

COMMON WRONG ANSWERS: Treating contextual inquiry as usability testing or a demo of your prototype. Expecting statistically significant sample sizes or quantitative metrics from a small-sample qualitative method. Dismissing the findings because they conflict with your existing architecture or because users are doing it wrong. Saying the goal is to validate a specific design rather than to discover unknown work practices. Claiming it is purely a UX activity with no engineering relevance. Another red flag is insisting that a product manager can just write down the requirements for you.

LIKELY FOLLOW-UPS: How would you translate a finding like users habitually saving every few seconds into concrete engineering requirements. When would you recommend against using contextual inquiry. How do you balance small-sample qualitative depth with the need to prioritize engineering work for a large user base. What would you do if the inquiry revealed that the real user workflow happens mostly in a competitor tool or spreadsheet. An interviewer might also ask how you would convince a skeptical engineering manager to budget time for this research.

ONE CONCRETE EXAMPLE: In an auto-insurance data-entry tool, specialists reported simply copying spreadsheet data into the software. During contextual inquiry, researchers observed that the specialists were also cross-referencing a second application to fetch missing correlating data and were hitting Save after every manual entry because they did not trust the autosave feature. These invisible steps led to design changes that addressed integration needs and trust issues. As an engineer witnessing this, you might realize the system needs a bulk import API, better cross-reference linking, and clearer save-state feedback.

Source: nngroup.com