|Announced on||Rollout starts|
|May 1, 2023||April 28, 2023|
We’re thrilled to announce the newest leap Guide is taking towards intelligent help centers, semantic search. Unlike search methods that find literal matches to keywords, semantic search considers the meaning of search queries to surface what the user is truly looking for within the top search results. This helps end users and agents to locate content without prior knowledge of the exact keywords to use. This announcement contains the following topics:
- What is changing?
- Why is Zendesk making this change?
- How to check if your help center is enabled for semantic search?
- What do I need to do?
What is changing?
Guide search has historically relied on keyword matches between queries and content (for example, articles, community posts, or external records). However, it didn’t capture the “meaning,” or semantics, of a query. Semantic search, on the other hand, deploys natural language processing to understand the content of the search query and content.
For example, with semantic search, end users and agents can ask natural language questions instead of worrying about which keywords to use in order to get the best results. New agents can ask “how do I start using Guide?” instead of trying to think of the best keywords to use or the sequence of keywords to use in their query. With semantic search, they can enter their question in the way that makes sense to them.
When semantic search is used, the search engine boosts the relevance of semantically better matches. This pushes the most relevant content, based on intent and context of the search query, to the top of the search results.
Why is Zendesk making this change?
We want to make it intuitive to find useful knowledge for anyone, and recognize that it’s hard to do with a keyword based search only. Knowledge teams create amazing help center content today, and this first release of semantic search is the beginning of a journey to use artificial intelligence (AI) to truly understand what users are searching for.
By improving the search relevance overall, and specifically ranking the most relevant results on top, semantic search produces the following help center improvements:
- End users can find the information they are looking for more easily, increasing the deflection power of your help center.
- Agents can be more efficient as they can find answers more quickly.
How to check if your help center is enabled for semantic search?
We’ve begun our semantic search implementation for Guide customers with English help centers. Approximately 50% of all help centers are currently supported by semantic search. As the rollout continues, we will continue to expand the feature to all Guide customers.
- Perform a search in your help center.
- On the search results page, open Inspect element in your browser.
For example, in Chrome, right-click the search results page, select Inspect from the menu, then click the Elements tab.
- Select the <head> tag, then examine the
If the element is set to "false" (as shown in the image below), semantic search is not yet enabled in your help center. If the element is set to "true", then semantic search is enabled.
Why is semantic search not enabled for my help center?
Semantic search is a search enhanced by using AI to improve search results. Today Zendesk Guide has three types of content, namely articles, community posts, and external content, across many different languages. For each content type and language, we have to identify the best AI strategy to achieve optimal search results and validate that this truly works better than the current search experience.
So far we have optimized our semantic search model for English and for articles since English is the language that sees by far the most searches and articles is the content type all help centers have.
This means that semantic search is currently available in English help centers (all English variants) and only in help centers that use the two-pane search results list. The two-pane search results list displays article search results side-by-side with the community post results. The list of article search results is powered by semantic search.
We are currently testing semantic search for community posts in English. Our upcoming release will offer semantic search for community posts in English. In addition to fully covering the content types for the two-pane search results, this will power the unified search SERP (one list of search results) on help centers that do not use federated search and thus have external content. After community posts, we will expand the model to support external content and more languages by search volume and extend is to more search interfaces like the knowledge pannel, the unified search API, instant search, suggested articles etc.
Two pane search vs unified search SERPs
What do I need to do?
This implementation is automatic and requires no additional work on the customer end.
We will also expand semantic search to cover more languages and areas of the help center and improve the semantic understanding of the language model. Expanding the feature will take a longer period, but we'll announce updates to the feature as they happen.
To help ensure that the upcoming improvements tackle the issues your help center visitors are facing today, we invite you to share examples of unsuccessful searches through this form. This feedback will guide our product team, and will also be used to evaluate the search performance.
Use this Feedback form to share your feedback with us.
If you’re interested in how semantic search works, read more in About semantic search and how it works.