Merchandising and promoting
On this page
Merchandising is about promoting records and filters. The items you promote may not always be the most relevant records, and they might not match the query as well as some other records. However, their presence at the top of the results serves your business objectives.
Merchandising is done for many reasons: To offer recommendations to your users; to highlight your inventory’s best products; to follow trends; to improve bounce rates. Controlling results in these ways helps encourage your users to stay longer on your website, to browse deeper into your catalog, and to purchase items that align best with your business strategy.
Algolia offers two merchandising tools.
- Query Rules (Algolia’s main merchandising tool)
- Optional Filters
Optional filters are available on all plans. Query Rules are available on the Enterprise plans only. See our pricing page.
Merchandising is Powered by Query Rules
Query Rules allows you to target specific search terms and alter the way the Algolia engine would normally treat those terms. Essentially, with Query Rules, you single-out some queries for special treatment, altering their normal behavior. It can be used for merchandising or promoting items as well as for understanding and reacting better to your end-users’ intentions. On this page, we focus on merchandising and promoting. You can go here for intent detection.
We highlight here the most common merchandising techniques using Query Rules.
Hit promotion can be expressed in the following way: “I want Record X to appear first in the results for a specific query”. This is as unambiguous as it gets. A bookstore wants to recommend a Harry Potter Box Set whenever the words “Harry Potter” appear in the search.
Another example of promoting hits is with managing new releases - say, the newest iPhone. You’ve placed “best-selling items” at the top of your search results by using Custom Ranking. Unfortunately, this does not work well with new releases, which have no sales yet. Where do new releases go, then? Probably on the last page. With Query Rules, you can force new releases at the top of the results list without changing your custom ranking.
You can also hide hits. Your index may contain records relevant only to a specific search. Let’s say your index contains “banana t-shirts”. If someone types in “banana”, you don’t necessarily want to show t-shirts. Hiding hits allows you to remove all banana t-shirts from the “banana” query.
An extension to promoting hits is to display a banner that can promote an item, or a group of products, or a business-wide sale, and so on. Using the same logic - that is, parsing a query, looking for specific triggering keywords - you can create a rule that contains a banner - such as the text of the banner, its HTML and CSS, and any associated images.
Dynamically Promoting Filters
You can also promote filters with query rules. This functionality is treated in more depth as part of Query Rules’ intent detection functionality, but it’s worth noting it here as part of an overall merchandising strategy. If you detect that a user wants to see, and could benefit from, a particularly important category of products for your business, you can promote those products using dynamically generated filters. You can also dynamically generate optional filters, which are discussed immediately below.
Using Optional Filters
Optional filters enables you to favor some filters over others. They behave like regular filters, except that with optional filtering, results that do not match a filter are not excluded altogether, they’re only ranked lower in the result set. For example, if you know a user is interested in the brand “samsung”, you can promote that brand with optional filters: when that user types in “phone”, all phones from Samsung will be returned first, followed by all other phones.
Combining Multiple Query Rules
All of the above merchandising techniques can be used individually or in combination on the same query. Another way of putting this, one rule/condition can have many consequences. For example, if a user types in “Harry Potter”, you can emphasize films over books, show the most recent Harry Potter box set, and also add filters for JK Rowling, Children’s novels, or Fantasy genre, all based on one condition (“Harry Potter”).