Duolingo Metrics & Unit Economics interview questions

AI features are unusually good at producing engagement that is actually friction. The question tests whether you can tell the two apart in the data.

A principal-level question because the cost of serving a user is the same in Mumbai and Minneapolis while the price you can charge is not. The free tier is an acquisition budget, and the paid tier needs a model mix by market.

Turn a registered-user count into active usage, model calls and billable tokens. Compare an uncached baseline with a stated caching scenario without inventing a vendor quote.

An estimation question that tests commercial judgment. The inference bill is where the arithmetic starts; the number that decides the answer is somewhere else.

A vocabulary question with a sharp edge. The cheapest model per token is regularly the most expensive per finished job.

Usage counts are the trap. The answer that scores connects quality, behaviour, cost and risk to one business outcome, and says what evidence would change the primary metric.

The cheap option per conversation is the expensive one per customer, until you read how the vendor defines 'resolved'. Two calculations, and the second one is where the offer is decided.

Candidates jump to 'switch to a cheaper model'. The interviewer wants to see you find out why cost per user rose before choosing a lever, then pull the levers in the order that protects quality.

A vocabulary question about speed. The right metric depends on whether the user is watching text stream in or waiting for a finished result.

More AI-generated code can mean more code to review and maintain. Choose a metric that moves when developers deliver useful, reliable work.

A margin question with a constraint added halfway through. The average user is not the problem; the distribution is, and a flat price hides it until the model gets more expensive.

Survey says everyone saves five hours a week. Finance does not believe it, and they are right not to. The answer is to measure a few workflows properly instead of the whole company vaguely.

AI metrics have unusually soft edges: what counts as an AI product, a resolution or an active user can widen without anyone announcing it. This question checks whether you audit a good number as hard as a bad one.

The candidate who says 'use a faster model' in the first sentence has already lost points. Decompose the budget, find the biggest bar, and check whether two seconds is the right target at all.

Routing is one of the biggest cost levers in AI products, and one of the easiest to get subtly wrong. The interviewer wants the savings arithmetic and the misrouting risk in the same answer.

The most expensive AI failures are the confident wrong answers that users accept. This question tests whether you can find their footprints in behaviour data.

Caching is the cheapest large saving in most AI products, and a single line of personalisation at the top of the prompt can turn it into a 22% cost increase. The arithmetic is short; the layout insight is what scores.

A portfolio and stakeholder question. Across-the-board cuts are the easy answer and the wrong one. The strong answer finds where spend creates value, cuts waste first, and gives the CFO a number they can hold you to.