Propose two research methods to investigate checkout abandonment

WHAT THIS TESTS: The interviewer wants to see if you understand research triangulation. Checkout abandonment is a messy problem with multiple causes, price sensitivity, usability failures, technical bugs, and trust issues. Picking two methods that yield the same data type shows a junior mindset. The question tests whether you can map method to insight: one method to locate the leak and quantify it, another to explain the human reasoning behind it.

A GOOD ANSWER COVERS: First, a quantitative behavioral method such as funnel analytics or session replay. This tells you exactly where users drop off, whether it is at shipping, payment, or account creation, and whether the pattern is device-specific. Second, a qualitative observational method such as moderated usability testing. This reveals why users abandon: Baymard research shows users often leave because they cannot complete fields, encounter unexpected costs, or become infuriated by forced account creation. The two methods work in sequence; analytics narrows the scope, and usability testing generates hypotheses about cognitive load and emotional response. A senior candidate also mentions sequencing, running the quantitative sweep first to prioritize which checkout steps to observe in the qualitative sessions.

COMMON WRONG ANSWERS: Proposing two self-report methods like exit surveys and user interviews. Both capture what users say, not what they do, and self-reported data about checkout is notoriously unreliable because users post-rationalize leaving. Suggesting A/B testing as a research method; testing is for validation, not discovery. Proposing analytics without a follow-up qualitative layer, which leaves you with a dashboard but no actionable redesign target. Ignoring the distinction between checkout flow issues and general price sensitivity; Baymard data shows checkout design itself is frequently the sole cause of abandonment, not just the product cost.

LIKELY FOLLOW-UPS: How would you recruit participants for the usability tests? How do you distinguish between a usability issue and a pricing issue? What would you do if analytics showed a 40% drop at the shipping step but users in testing said the price was fine? How would you prioritize fixes if you have limited engineering resources?

ONE CONCRETE EXAMPLE: Say analytics show a 35% drop-off at the shipping information step. Session replay shows users hesitating and returning to the cart page. You then run moderated usability tests with five participants and observe three of them struggling to find the apartment number field, causing validation errors. One participant says, "I am worried my order will ship to the wrong place," and abandons. The quantitative data localized the problem; the qualitative data revealed it was a form layout issue, not a shipping cost surprise. You now have a specific fix: expose the secondary address line by default.

Source: Baymard Institute