Google Product Strategy for a Real-Time Mood-Detection API

You are a product manager at Google. Your team has built a mood API: a service that can tell, in real time, what mood any given person is in. Leadership wants to know what Google should do with it. Lay out the opportunities this capability opens up, decide which one Google should pursue first and for whom, and explain how you would measure success and manage the risks of a product that reads people's emotions. ```hint Sort ideas by who receives the mood The same mood reading means something very different depending on whether it goes to the person it describes, to someone they chose, to a Google product or to an outside company. Group your ideas that way before you rank them. ``` ### Clarifying Questions - How does the API infer mood: from a wearable's heart-rate signals, from voice or typing, from phone activity, or from something else, and how accurate is it? - Does "any person" mean any caller can query anyone, or only people who have opted in and granted access? - What does the API return: a handful of named states such as calm or stressed, a continuous score, or both, and with what confidence level? - Should Google use it only inside its own products, offer it to outside developers, or both? ### What a Strong Answer Covers - A goal tied to Google's mission and business, with an explicit position on consent and on who may read whose mood. - A structured map of opportunities (personal use, contacts the person chooses, Google's own surfaces, clinical partners, outside developers, advertisers, institutions) and the users each one serves. - Prioritization by user value, fit with Google's products, whether the model's accuracy is good enough for each use, and trust risk, ending in one concrete first product. - Success metrics centered on benefit to the person, with guardrails for false readings, misuse and privacy. - Clear limits on uses Google should refuse, plus mitigations for regulation, accuracy gaps across groups, security and reputational damage. ### Follow-up Questions - An ads team wants mood as a targeting signal and projects a large revenue gain. How do you respond? - The model turns out to be noticeably less accurate for some groups of people. What changes in your launch plan? - Would you open the API to outside developers at launch? If so, what would you require of them?