What instrumentation would you add to validate the user's 'job'?

What's really being asked

This question tests whether you understand that Jobs to Be Done is about the progress a user is trying to make in specific circumstances, not about feature popularity or user demographics. The interviewer wants to see if you can design telemetry that captures intent, struggle, and context at scale. Engineers often instrument for vanity metrics like button clicks; this probes whether you can build logging that reveals the functional, social, and emotional forces that pull users toward or push them away from using the export.

The full answer

A strong answer proposes event logging that maps to the full job timeline. First, capture pre-export context: what queries, filters, or date ranges led the user to the export button, because the job often begins before the feature is touched. Second, instrument the export flow itself with step-level abandonment tracking to identify where anxiety or confusion blocks progress. Third, log post-export behavior: whether the file is downloaded, opened, shared, or re-imported elsewhere, since the job is not done when the file leaves the system. Fourth, detect workarounds like copy-pasting data, printing to PDF, or taking screenshots, which reveal the real job users are hiring the product to do when the official feature fails them. Fifth, add lightweight qualitative signals such as rage clicks, long pauses, or help-menu searches during the export workflow.

The mistakes people make

Red flags include proposing only aggregate usage counts or demographic segmentation, which the Christensen Institute explicitly contrasts with Jobs Theory. Another mistake is suggesting an exit survey without any behavioral instrumentation, since that relies on recall rather than observed behavior. Proposing to track only successful exports misses the struggling moments where the real job insight lives. Similarly, suggesting A/B tests of button colors or tooltip copy before understanding the underlying progress the user seeks puts tactics ahead of diagnosis.

What usually comes next

The interviewer may ask how you would prioritize these instrumentation points given engineering constraints, or how you would distinguish between a user who exported data for compliance versus analysis without asking them directly. They might also probe how you would validate that a detected workaround actually maps to the same job, or how you would instrument for social and emotional forces like the need to look competent when sharing a report with leadership.

A concrete example

Suppose users frequently apply a complex filter, reach the export preview, and then abandon the flow. Instrumentation shows they immediately open a spreadsheet tool and paste data manually. The hypothesis is that their job is not to own a CSV file but to answer a specific question before a meeting. To test this, you instrument for preview-stage behavior, track whether exported files are ever opened, and detect manual copy events. If the data shows preview abandonment correlates with same-day manual spreadsheet work, you have evidence the job is rapid insight, not file ownership.