If you’re already a Hitchhiker, log in to access this content.
Search Quality QA| Hitchhikers Platform
What You’ll Learn
In this section, you will learn:
- An overview of Search Quality QA
- The 3 types of search quality issues
What is Search Quality QA?
The second phase of QA focuses on the actual search quality of your Answers experience. You will be confirming that the most relevant and appropriate information stored within the Knowledge Graph is being correctly returned for a given query. In most cases we need to lean on the client and their knowledge of the business to provide feedback and ensure the quality of the search results.
When assessing the search quality of an Answers experience, you’ll want to run through 20-40 queries and ensure the relevant results are being returned. We recommend identifying or creating a list of ‘top priority queries’ for the brand, which will be a great starting point to QA your experience. For example, if you were searching on the Turtlehead Tacos website these are some of the queries you might be asking: “Restaurants near me”, “How many calories are in a breakfast burrito?”, “What allergens are in a fish taco?”, etc.
In an ideal world, each of the queries you assess will return the expected results without having to make a single adjustment. However, this is rarely the case and oftentimes there are adjustments that you will want to make upon reviewing the experience in more detail. No need to worry! This is all part of QA.
What are the 3 types of search quality issues?
When assessing the results for any given query there are three main buckets that categorize the issue or help determine why the relevant results are not being returned.
- Knowledge Graph Issue: There is no content in the Knowledge Graph that addresses the query.
- Configuration Issue: Searchable fields, vertical intents, synonyms, hardcoded prompts, or other backend configuration options need to be adjusted to return the relevant entities.
- Algorithm Issue: Although the Knowledge Graph has relevant data and the configuration is as optimized as possible, the Answers algorithm is not able to correctly match or rank the content.
Not sure what might be causing the issue you’re seeing? Post in the Community to ask your peers and Yext experts.
Once you can identify which issue types you’re facing, you’ll be more equipped to debug the issue — which you will learn about in more detail in the next module.