Reimagining search: a product vision for how people will find information

You are a product manager asked to picture how people will look for information in the future and to design the search product that serves them. Explain what is changing that makes search as it works today fall short, decide whose needs you are designing for, describe the product experience and what you would build first, and say how you would know it is better than today's search. ```hint Trace a session to the end Follow one person's search from the first query to the moment their need is actually met, and note every step where they still do the work themselves. ``` ### Clarifying Questions - How far ahead is "the future": a product we could start building now, or a long-range vision that assumes technology not yet available? - Is this the next version of Google Search for its existing users, or a new product that may live on different devices and surfaces? - What should the design optimize for: needs met faster, more usage, or the health of the ads business, and may it change how search makes money? - Is the scope the open web only, or may search also draw on a person's own email, files and photos with their permission? ### What a Strong Answer Covers - A point of view on the technology and behavior shifts that change what people need from search, with a stated time horizon - Kinds of searchers or search tasks compared, with one chosen and the reason - The chosen group's pain points with today's search, ranked - A concrete product experience, with ideas prioritized against explicit criteria and a first release defined - Success metrics that show needs being met, guardrails, and the risks to trust, revenue and the sites search depends on ### Follow-up Questions - If the product answers questions directly, why would publishers keep letting it read their pages, and how would you keep that ecosystem healthy? - How would advertising, or another revenue model, fit your design without eroding trust in the answers? - How would you detect and contain an answer that is confidently wrong when it is shown to millions of people?