Defining Relevance

Relevance Is …

Intelligent Matching

Relevance makes sure that all records that match a query are found. However, what does it mean to “match a query”? Relevance is an intelligent matching that takes into account typo tolerance, partial word matching, spatial distance between matching words, the number of attributes that match, synonyms and query rules, natural language characteristics like stop words and plurals, geolocation, and many other intuitive aspects of what people would expect from search, especially in the Google era.

Finding records is only part of the story. Once you have the “right” records, another aspect that comes into play is how the records are ordered, or ranked.

Ranking - Putting the Best at the Top

Relevance is also ranking. It’s how you order the records that a search returns so that the most accurate results appear earliest (on the first couple pages), while less accurate results appear later.

When we speak about ranking we use the terms textual relevance and business relevance. Textual relevance is about the strength of a match. Records that textually match better than others will be ranked higher. However, if some records have equal textual relevance, that is, they are tied, then Algolia offers you a way to break the tie and control the order with business relevant metrics. This is called business relevance (or custom ranking).

Putting them together - Matching and Ranking

Relevance is about finding all the records that match a search and then ordering them so that the most accurate results are the most visible. If the best match is lost on the last page, or in the middle of hundreds of pages, you’ve failed to achieve relevance. In this sense, while finding the right records is important, if the user doesn’t see the best, you’ve failed to achieve relevance.

So relevance is a balance between finding records and putting them in the best order. But does it stop there?

Promoting your business

Relevance is also about promoting your business and creating meaningful context, using such tools as custom ranking, merchandising, personalization, and analytics. With these tools, relevance is about creating results that best showcase and sell your products.

Meeting User Expectations

These tools also help match your end user’s expectations. Relevance is about user expectation, giving the user what they are looking for. When a search matches the user’s intent, you’ve achieved relevance.

Discovery

And it’s about discovery as well. Users are not always looking for a specific item. They are browsing, defining what they need with each new query. In this sense, relevance helps a user decide what they want. Here is where the speed of the search and the user experience on the front end add to if not fully define the meaning of relevance. This is the aspect of conversational search that is central to providing relevance.

UI/UX

Finally, relevance is a UI/UX concern. The speed of the search, the highlighting of key terms in each result, faceting and filtering, pagination and infinite scrolling - in a word, relevance is the overall user experience that allows users to navigate through their results to find, discover, and refine their search.

The whole is greater than the sum of its parts

Nonetheless, all of this presupposes finding and ordering (ranking) the best records. With these twin criteria taken care of, a foundation is built upon which intent, expectation, discovery, promoting and showcasing your business, and good UI/UX can thrive.