Start your content search
9 min read

Related Tags overview

OK. I want to find a dress… Hmm, long?... Er, short?... Maybe midi? Yeah! Definitely, I’ll go for midi!

Related Tags are descriptive keywords that complement and enhance the search query to make it easier for shoppers to find what they’re looking for with just one click. They’re closely related to the search query, fine-tuning the search to get highly relevant results. They only appear after the search process is completed, and the results are displayed.

For example, people usually start searching using simple, broad term queries, such as “beer". When they get loads of results, they realize that what they’re really searching for is a specific kind of beer. That’s when Related Tags come into play, providing extra and relevant suggestions for a more precise search, like pilsen, IPA, non-alcoholic, or light.

With Empathy Platform, you can show organic, curated, and AI Related Tags in your commerce store:

  • Organic Related Tags: These are generated automatically based on collective behavioral information, offering related keywords to the searches that are used by other shoppers for a more accurate search.
  • Curated Related Tags: They let you curate the search experience and suggest related search terms that shoppers might like to use to complement their initial search to drive sales, for example, as part of a Christmas campaign. As search manager, you may want to guide shoppers towards party dresses by associating sequined or festive as Related Tags for the query "dresses", in addition to the organic Related Tags party, black, or long.
  • AI Related Tags: They're automatically generated using AI mechanisms from catalog data. They can extend Related Tags coverage for instances with low organic traffic and address cold start challenges effectively. AI Related Tags can display together with organic and curated Related Tags as long as there's enough traffic volume for the initial search query.
How Related Tags types work together

When a shopper performs a search, all three types of Related Tags can display together in the same list, in the following order: curated, organic, and synthetic. The type of Related Tags that appear in your commerce store depends on what's available for that given query.

As curated Related Tags are keywords you willingly add to complement the initial query to promote products, anticipate trends, and more, they display in the first place in your commerce store to guide shoppers through the search experience you want based on your business needs. Organic Related Tags are placed after curated ones and ordered following the preferences of other shoppers' experiences. AI Related Tags always appear at the end of the list.

Tip

By default, you can't distinguised between one type from another when Related Tags display in your commerce store. However, if you want shoppers to know which of them have been willfully created as a sign of transparency, you could differentiate Related Tags somehow with an icon, for example. Check the options available in the Related Tags UI component.

You can use the Related Tags management tool in the Empathy Platform Playboard to review the organic and AI Related Tags generated for a query, create new curated Related Tags that respond to marketing strategies, change the order in which the curated Related Tags appear, and choose to show or hide Related Tags.

interact

Want to learn more about how you can tailor the search experience for Related Tags? Read how to manage Related Tags in the Empathy Platform Playboard.

Brand value in personal care and beauty retail

See how curated Related Tags can benefit your shoppers and your business

Spot the difference

Don’t confuse Related Tags with Facets and Filters. Facets and Filters narrow down the search results to pinpoint products from a wide range of hits, while Related Tags complement the original search query, launching a new search that can bring up a new set of results or refine the existing ones.

For example, you can search for the term “dress”, then select the “midi” related tag to refine the initial search and get more relevant results. Finally, you can select the “XS” filter from the Size category to filter and display all extra-small midi dresses.

Likewise, try to not mix it up with Query Suggestions since they refine the query while you’re typing it before launching the initial search. Bear in mind, however, sometimes the suggestions displayed for Related Tags and Query Suggestions are the same.

You can identify Related Tags as they look like actual tags or labels located beneath the search box, once a search is launched and the results are displayed.

note

Remember that other types of suggestions can display hints based on search history or even trending products.

  • Sharpen query formulation. Educate shoppers on how to make better queries.
  • Speed up query entry, especially on mobile devices.
  • Improve product findability. Guide shoppers to the right products.
  • Enhance the shopper experience. Amaze shoppers with accurrate searches for what they’re looking for.
  • Improve search performance. More specific queries means more accurate results.

Each type of Related Tags follows a different generation path in the background to display in your search experience.

Organic Related Tags are generated by shoppers' actions in the search UI. As shoppers perform different searches and actions, behavioral information (queries, clicks, add-to-cart, etc.) is collected using the Tagging microservices and processed in the Query Signals batch process to extract the most relevant data for a specific query.

The information is then analyzed to select the appropriate candidate tags related to a specific query. So, from queries such as "long dress", "midi dress", or "maxi dress", long, midi, and maxi are identified as the Related Tags for the query "dress".

These candidate Related Tags are then reviewed to check whether they are valid according to different filters (i.e. blacklists, duplicity, etc.), resulting in a list of final Related Tags that is ready to go.

Curated Related Tags are created manually by the commerce merchandiser in the Playboard and sent to the Beacon microservice to be stored with the organic Related Tags. When the shopper performs a search in the search UI, Empathy Platform Interface X calls the Beacon API to retrieve a list of organic and curated Related Tags associated with the query, displaying them in the commerce store after the initial search.

interact

Want to learn more about how you can tailor the search experience for Related Tags? Read Manage Related Tags to learn how create curated Related Tags in the Empathy Platform Playboard.

AI Related Tags are generated using the same underlying AI mechanisms than AI Next Queries. AI Related Tags are synthetically generated using an LLM. The process takes the store's top search queries, sourced from the Empathize batch process output, as its starting point. For each query, the workflow retrieves its search results and feeds them to the LLM together with the query and a prompt describing the generation task. These results give the LLM context about the store's catalog, so it identify semantically related search terms and generates short refinement tags, including ingredients, attributes, brands, categories, and turning them into Related Tags candidates.

The LLM's output then goes through a review process before it reaches shoppers. It filters out duplicates, near-duplicates, and irrelevant terms. It then keeps only candidates that are a direct and meaningful refinement of the original query, meaning they contain the original search term along with something more specific. So, for a query like "dress summer", summer is identified as a related tag for the query "dress". Suggested candidates are also checked against actual search results, and any suggestion that returns no results gets discarded. The remaining Related Tags are also normalized (for example, converted to lowercase and depluralized) to ensure consistent matching with how shoppers search.

Once validated and normalized, the Related Tags are loaded into the Beacon microservice and stored alongside the organic and curated ones.

When the shopper performs a search in the search UI, Empathy Platform Interface X calls the Beacon API to retrieve all Related Tags available for that given query regardless of the type.

note

AI Related Tags are powered by Empathy.AI (opens new window). The generation workflow runs on private, sustainable, self-hosted infrastructure, so no query, catalog, or shopper data gets shared with third parties, guaranteeing data sovereignty and operational resilience.

interact

Explore the interactive map to see how Related Tags relates to the other Empathy Platform features and microservices.

Cookie-Less