Example: Restaurants
Restaurants on a restaurant-review site are defined by many categories: food quality, cost, service, cuisine type, location, and more.
Most restaurant sites offer extremely primitive search. Users can sort on one characteristic at a time – perhaps two – but usually no more than that. As a result the mechanism most users employ to find the restaurant they want is a simple keyword search.
The Discovery Engine allows users to drill more deeply into a dataset of restaurants. A user might start by selecting a cuisine-type and an area. Unlike other search engines, the Discovery Search Engine wouldn’t stop after displaying all the restaurants that match this request. It would continue the list of search results with restaurants that serve the kind of food the user wants and are near, but not in, the neighborhood he wants. It would also show a few restaurants that are in the right location but serve a slightly different type of food. Eventually, the user would start to see restaurants that are neither in the neighborhood he wants nor serving the kind of food he wants, but are close matches on both counts. This is a big improvement over showing a user a short list of perfect matches and forcing him to repeat his search again and again or leave the site in frustration.
If the user wants to narrow results further he might continue to drill down by adding a rating for food quality. He would then be shown restaurants in the location he wants, with the cuisine he wants, and the food quality he wants. Although few restaurants might match perfectly on all counts, the Discovery Search Engine would nonetheless show him a long list of good matches by displaying the restaurants that are the closest matches across all three categories.
If the user selects a specific restaurant, the Discovery Search Engine can also offer the user restaurants which are close matches to that restaurant across all categories: location, cuisine, food quality, cost, and anything else. This gives the user a whole new way to browse a selection of restaurants. He may start with a restaurant he knows and spot a restaurant he’s never seen before which seems as good or better than the one he originally wanted. In this way, the Discovery Search Engine does more than help people drill down on what they want. It allows them to discover things they love but never knew existed.
Other Examples
 |
 |
 |
 |
 |
 |