The Search Query Journey | Yext Hitchhikers Platform

What You’ll Learn

In this section, you will learn:

• What happens when a user runs a query on universal search
• What happens when a user runs a query on vertical search

Search Query Journey on Universal Search

In this module, you’ve learned a lot about the different components of Search — including:

• Differences between Universal and Vertical search
• How federated search works
• How the Knowledge Graph impacts both the Search Configuration and the Frontend
• How the algorithm uses data from the Knowledge Graph combined with rules from the Search Configuration to determine the best results
• How the algorithm then passes those results to the frontend for display

Let’s put it all together now and look at a real example. Let’s say you search on Yext Search for “search”. Go ahead and do it now by clicking here .

What are all of the things that are happening under the hood in the mere milliseconds it took for you to get an answer? Let’s take a closer look at what’s happening.

Note: if you haven’t completed the previous units in this module, we recommend reading them now before you continue.

Step One: User asks a question

We glossed over this by sending you a direct link to the search. However, the moment a user clicks into the search bar, Yext Search is hard at work. The search bar itself guides users to suggested queries using both hardcoded prompts and popular queries. You’ll learn more about both of these in the Query Suggestions module . Here, you are searching for the word “search”.

Step Two: Algorithm routes queries via Federated Search

Once a user types in their search and hits enter, the search query is then sent to Search — first stop, the Federated Search. The algorithm routes the query to the different verticals, including both those backed by Knowledge Graph entities and those backed by Third Parties, like Zendesk or Google Custom Search.

In this example, the router will send the query “search” to the many verticals defined in the Yext Experience, plus the Zendesk backend we use for our support articles and Google Custom Search which we use for our links backend.

Step Three: Algorithm does some heavy-lifting and evaluating

Based on the settings in the Search Configuration and the relevant data, the algorithm produces a set of results, if they exist, for each vertical search. For Knowledge Graph verticals, Search has a predefined index of entities and attributes based on the rules provided in the Search Configuration. For Third Party Verticals, Search calls the respective search endpoints using any of the business logic or access information specified in the Search Configuration.

In our example, each vertical attempts to return valid results for “search”. You’ll see that verticals like Locations came back with no results, but FAQs, Products and Webinars all came back with at least one result.

Step Four: Algorithm aggregates the results and sends them via the Search API

The algorithm not only returns which entities should be returned, but also the order of the results. Now that the Algorithm has the potential results for each vertical, the Algorithm has to put these together in a cohesive order and structure all of the data needed and sends it via the Search API.

Step Five: frontend assembles an answer for the user

The frontend combines:

• The components from the Search SDK
• The code in the Repository
• The results from the API to build the user-facing results.

Search Query Journey on Vertical Search

The search query journey looks almost identical in Vertical search as it does on Universal, except the algorithm only has to search one vertical.