Global relevance configuration

Tailor user experience to improve business metrics by optimizing Search results page merchandising strategy


Why should you do it?

Relevance defines every aspect of the Algolia search experience: from how you structure your data, to how the engine queries your indices and orders the results, to what your users see as they type. A thoughtfully designed and well configured relevance strategy serves as a foundation for a successful merchandising strategy. You will gain efficiency and scalability in your day-to-day operations, fast and easy merchandising strategy launch and avoid unnecessary complexity and mistakes in your implementation. 

The initial setup

The initial setup of global relevance and merchandising strategy includes both steps performed by the business teams and the developer teams. 

To ensure that you are following the best implementation practices, refer to the Checklist for Going Live.

Global relevance vs merchandising strategy

As a first step of the merchandising strategy, you will need to configure relevance, to get the best results in all situations. To set up a proper merchandising strategy, we begin with a global relevance set up first. The main difference between the two is that global relevance impacts all the results, while the merchandising is applied to selected hand-picked sets of results only. An example of a well planned relevance configuration would be the global relevance strategy matching 80% of use cases, while the remaining 20% are adjusted with merchandising. Each eCommerce business might have slightly different needs, therefore the split in the impact on search results is likely to be adjusted for a specific use case. (Note: The 80/20 split is not indicative of a great balance, but rather an idea to get the reasoning behind global relevance vs merchandising).

When you design a search experience for your users/customers, there are three types of relevance you should be interested in, that need to be in balance in order to achieve an effective merchandising strategy.

Textual, business and contextual relevance

One the one hand, is the textual relevance, which will pull up results that match what the user is looking for; on the other hand, business relevance allows you to push results that you want your users to see for various reasons such as popularity, higher sale margin, promotional campaign, purchase history. The last one is the contextual relevance, which is based on the users’ context and interaction with the UI/UX.

There are several essential steps that let you configure the engine precisely to find the perfect balance between textual, business and contextual relevance and address the challenges set by both product and marketing teams successfully. 

To better understand the concepts and gain expertise in relevance strategy, please refer to the corresponding Algolia Academy section on Concepts of Relevance and Merchandising with Algolia. [link TBD]

For a step-by-step instruction on setting up the initial configurations in the Algolia dashboard, please refer to Business User Guide to Relevance and Merchandising in Algolia Documentation.

Multi-store and Multi-language support

In the case of a multi-store and multi-language ecommerce platform, there might be a need to adjust language settings and define synonyms that might be very specific or unique to a particular geographic location or language. 

Adjusting language settings allows you to provide a multi-language support for your store, to ensure all of your users are getting the same level of product search and  discovery experience. The default settings are to use all dictionaries. This will result in such anomalies as applying French spelling features to English. For dictionary-based settings, like typo tolerance, stop words, plurals, and others, tell the engine which languages you want these settings to use as their logical bases.

Defining synonyms is beneficial since users don't always use the same vocabulary as ours. When site users believe A and B are the same, they would expect the same search results for both of them. Synonyms will help you make sure user query words match your data in your existing attributes. Adding keywords in a new searchable attribute can have a stronger impact on relevance.

For a step-by-step instruction on setting up the initial configurations in the Algolia dashboard, please refer to Business User Guide to Relevance and Merchandising in Algolia Documentation.

Step-by-step instructions>Initial configurations

Takeaway 1


Takeaway 2


Going further

Additional Resources


Business User Guide to Relevance and Merchandising


Adding Synonyms

Algolia Academy

Configure Relevance with the Dashboard