Example: Wine
Bottles of wine on a wine-selling site are defined by many categories: color, region, rating, year, price, mouth-feel and more.
Today, most wine sites offer extremely primitive search. As a result the mechanism that most users employ to find the wine they want is a simple keyword search.
The Discovery Engine allows users to drill more deeply into a dataset. A customer might start by looking for a cabernet from California. After seeing the results he might narrow them by adding a price, year or rating.
Rather than tell a customer that there are no wines that precisely match his search, the Discovery Search Engine will offer users the closest matches in inventory. By always giving the customer a good selection of items that closely match what he wants, the Discovery Search Engine increases the chance that he will buy.
If the customer finds a wine he likes, the Discovery Search Engine can also show him the most similar wines in the system across all categories: color, region, year, rating, price and more. The user can then select one of these wines, and will in turn be shown still more similar wines. In this way the customer sees inventory than he otherwise would, all of which is related to an item he’s already interested in. This increases the chance that he will buy and the potential size of his order.
If certain wines are pushed by a vendor, these wines can be given a degree of “buoyancy” by the Discovery Search Engine. Thus, if a customer is looking for a California cabernet in a certain price range and one bottle, which would otherwise be ranked 12th out of 15,000 in response to his search, is on a list of wines to push, that bottle might instead show up as the 8th best match to his search request. This allows vendors to sell through selected product while still giving customers extraordinarily precise matches to what they’re searching for.
Other Examples
 |
 |
 |
 |
 |
 |