1. Product sense - design a travel product Awesome. I love to travel. If I were building a travel product I'd probably start by thinking about a variety of use cases I've experienced to brain storm out areas where there could be an opportunity to create value in the already competitive market. I'd also look at what's out there as a benchmark. Use Cases: 1 Convenient and easy way to book travel needs (flights, car hotel) - Data aggregation - Seamless checkout 2 Reason for booking travel - Leisure - Business 3 Travel flexibility (fixed dates, open ended, cheapest flights anywhere) In looking at those use cases, there are a lot of ways we can go... so next I'd probably think about where to focus efforts. 1. There are a ton of flight aggregators out there with established brands, which would be hard to compete with. I'm not sure the world needs another kayak or google flights right now. 2. Understanding why someone is booking travel and where they are going could be a compelling way to enhance their experience. Business travel is usually booked after someone already has an agenda. Leisure travel is usually more flexible, so there could be a compelling way for us to help users design an experience once we know where they are heading. Based on this, I'd probably select this as think the impact is pretty large and there is some market validation out there from companies like TripAdvisor who have curated thousands of reviews from travelers. 3. The majority of people have relatively fixed travel dates, so in terms of impact, I think I'd rule this one out. In a scenario where I had more time, I would go talk to different people about the idea for some external validation. I'd also do more market research, bu we ain't go no time for that... Now that I have a north star I'd establish a couple goals, which might look something like: ** Helping travelers discover the world and connect with those around them ** Solutions: 1. User tells us where they are going. We then look at a variety of data sets to put together an itinerary that recommends the best things for them to do. 2. We could also profile the user to understand their likes and interests and use that to inform the itinerary. --> E.g. someone who doesn't drink may not want to visit the guiness factory in Dublin. add ons -- fb data, friends' favorite things --> network effect Actual implementation, I think we could start by creating a simple form where users give us a bunch of data. from there we'd manually create an itinerary for them. To determine success, we'd look at how many of rec'd activities the users purchased. We would make a commission on activities booked. As we validated the MPV and had a human algorithm that delighted customers we could then leverage a number of data sources into a data warehouse and write the algorithm to build itineraries. Growth Hacking, we could have the algorithm design thousands of itineraries for various places to drive organic SEO. KPIs: - % of form completion (are we asking for too much or too little? find the sweet spot) - Percentage of activities booked - Net promoter score