The Ranking Diagnostics tool offers you a comprehensive insight into Bloomreach's search retrieval and ranking process. It equips you with essential data to effectively troubleshoot search issues for any query/category. The tool runs a specific search and returns the search results to you with data that indicates how each piece of product information contributes to the ranking of products in search results.
There may be instances where you want to know why a particular product is part of the recall or why Product A ranks higher than Product B — Diagnostics helps you answer such questions with ease. Not only does it make the inner workings of our machine-learning models more transparent, but it also empowers you with greater independence to diagnose and debug recall/ranking issues.
Use the information to better understand the drivers behind a particular result set in Bloomreach Personalization, and determine whether any adjustments must be made. The insights gathered from the Ranking Diagnostics tool can help you determine whether a synonym needs to be removed, a particular product attribute in the product feed is incorrect, or some products need to be boosted or buried. You can then take action using the other Dashboard tools.
We recommend that you read the following guides to learn more about how the Bloomreach Engine processes search queries and the importance of product Recall and Ranking before debugging your searches:
Here’s what you do to analyze how products in search results are ranked:
- Navigate to Search & Merchandising > Operational tools > Ranking diagnostics in your Bloomreach Dashboard.
- Run the search that you're interested in.
- Review the results displayed by the Ranking Diagnostics tool to understand why particular products are showing up in search results and why they are in the sequence displayed.
After you run a query in the Ranking Diagnostic tool, the tool displays information that you can review to determine why a product is ranked a certain way and if you want to make a change.
Diagnostic information is divided into two parts. The top part of the page gives you information on how the query was interpreted and processed and the synonyms related to the entire collection of search results. It also points out other factors that might affect the query processing. The bottom part of the page displays the search results with a ranking diagnostics summary for each product. You can click on a product to view more details.
Here's an example of the top part, showing you the query "blue living room couch" and the different types of synonyms related to that query:
After you run a search, the bottom part of the page gives you a brief summary of information about each product in the search results. Here's an example with the "blue living room couch" query again:
You can click on a product to open a detailed summary of how Bloomreach used information about that product when the search results were ranked. Use this information to analyze why that particular product is part of the search results.
Check the Matched Terms section to see a list of standard fields and a list of terms in the query that matched these fields. In the example, the Festive Crush dress matched "dress" and "red."
Standard fields include product title, color, category, and brand. Any information about a product in your feed that isn't associated with the standard fields is included in the Matched Terms within the Other Searchable Field section.
Use the back button at the top left, next to the product PID to go back to view all the results.
For each product returned by Bloomreach, there's a subset of the following signals that were used to derive its position in the search results. Each signal has a normalized score from 0 to 100 that is calculated specifically for that query and result set. A score of 100.0 indicates that the product had the highest score among the result set for that particular metric.
- Total Score: the overall score Bloomreach gives the product for the specified search query. This score determines the position of the product among the result set so the sequence of the results is always in descending order from 100.0.
- Performance: this score is an aggregated metric that examines activities like click-through rates, add-to-cart events, and conversions of this product associated with the specified search term.
- Relevance: this score is derived based on how related the search term is to the attributes provided to Bloomreach in the product catalog feed, such as title, category, brand, and other fields that you include in your feed.
In addition, you can click a specific product to see the following information:
- Total Revenue: revenue attributed to the product over the past 60 days.
- Top RPV: revenue attributed to the product for each product page view.
- Text matches: the fields that matched the various keywords in the search term, such as color, category, title, and others.
Updated 6 days ago