Business cases and product questions
For non-engineering roles, the technical round is a case. The method is the answer.
"Our delivery times have slipped fifteen percent in Jeddah. What would you do?" There is no right answer, and both of you know it. What is being assessed is whether you can decompose an ambiguous problem, decide what evidence would settle it, and reach a recommendation you can defend — under time pressure, with someone watching.
Do not start with solutions
The instinct is to start listing interventions: hire more drivers, change the routing, add a surcharge. This is the fastest way to fail the round. A candidate who proposes fixes before establishing the cause is telling the interviewer what they would do to a real business with real money.
A structure that holds up
- Clarify. What exactly slipped, over what period, measured how? Fifteen percent against what baseline — last year, last quarter, plan?
- Segment. Is it all of Jeddah or specific districts? All hours or peak only? New drivers or established ones? Almost every business problem collapses once you find the segment carrying it.
- Hypothesise. Name three or four plausible causes across different categories — supply, demand, operations, external — and say which you would test first and why.
- Decide what would falsify each. "If it's a driver-supply problem, I'd expect acceptance rates to be down and wait times worst at peak. If those look normal, it isn't supply."
- Recommend. Say what you would do given what you know now, what you would measure, and what would make you change your mind.
Use numbers, even rough ones
If they give you data, use it explicitly and say what it implies. If they do not, estimate out loud and label the estimate: "assume a driver does six deliveries in a shift — I'd want to check that, but if it's roughly right then we're short about forty drivers, not four hundred, and that's a very different intervention." Quantifying an assumption is worth more than avoiding one.
Say the trade-off you are accepting
Every recommendation costs something. Raising driver pay fixes supply and destroys unit economics; cutting the delivery radius fixes the metric and loses customers. Naming the cost before the interviewer names it is the clearest signal that you have made a decision rather than picked an idea.
Close cleanly
End with a recommendation in one sentence, the single number you would watch, and the thing you would check first if you had a week. Interviewers remember endings; candidates who trail off into "so, yeah, lots of options" lose a round they had otherwise argued well.
They are not buying your answer. They are buying the method that produced it, because that is what you will bring on Monday.