Part 1: Product Ideation and Hypotheses Can you come up with 1-3 hypotheses/ideas that could increase the % of shoppers who convert? Now, pick the hypothesis that you think is most impactful. Can you quantify the potential revenue impact or value of this change? You'll need to make some assumptions. If we build x feature, what 4 main KPIs would you look at in order to measure the health of feature x? Part 2: Data Analysis and Interpretation We tested 4 tip variants compared to the control. The variants appeared either pre- or post- booking and as variable amount options or a toggle. Our most important metric is total revenue per user. Which variant performed best? If you were leading this experiment and these were the findings, what would be your next step? Part 3: Modeling As the provider of the Price Freeze product, Hopper has full autonomy on the pricing, structure and terms of each Price Freeze. Imagine that as a Data Scientist, your task is to provide a framework for structuring/pricing Price Freezes dynamically in order to maximize revenue. What do you think are the important drivers of success that should be tuned in order to achieve your goal? Given the drivers you've listed above: how would you model net revenue as a function of those drivers? Now, imagine you start building this net revenue model so you can predict the best Price Freeze to offer the user. You have the ability to deploy one or many models into production and observe how users interact with the Price Freezes you offer them. How do you design the experiment? What are you trying to learn? What does success look like? How do you progress, given the various potential outcomes?