Netflix relies on recommendations to help users find content, and emphasizes personalized suggestions of shows and movies for users on the home screen. These recommendations are important because Netflix’s business model depends on watch time; they want users to stay and watch for as many hours as possible. So, their recommendation system helps retain users by reducing decision fatigue and automatically surfacing relevant titles that the user would stay to watch next based on their history, like/dislike, and watch time. Similarly, they’ll evaluate their own titles based on watch hours to see what users generally like in each region and generate new content on their end.
YouTube combines search discovery with recommendation algorithms. Often, users know what they want to search for, and YouTube needs to get them there. After the user is watching something, the algorithm generates recommendations to shift the users into a passive consumption loop through autoplay and sidebar suggestions. YouTube’s business model is driven by ad inventory and user attention, so the this optimizes for session length for ad exposure rather than completing a content or evaluating how much a user likes the content. They blend agency with recommendations: help get the users to a spot, then feed them more videos.
Airbnb focuses on filter-heavy browsing to help users navigate a marketplace of listings. Users’ living styles drastically vary, so with filtering, Airbnb needs to tailor the search experience to find users bookings that they’d want. They prioritize high user control; a user implements filtering to set location, price, amenities, and dates to get to listings. Airbnb emphasizes trust and decision confidence rather than anything time-related since they depend on booking conversions; they want to guide users to their best option as effectively as possible to encourage a transaction to take place.
