The Results Preview is the section where you can test your Search Engine to see the results and assess the configuration settings for your product feed. In this article, you will understand all aspects from Results Preview.
Find it in your Admin Panel > Search > Results Preview.
Results Preview Panel Overview
Once on the panel, you will see the search bar. Search for a specific product on it.
If you want to finish the search, a cross “x” appears on the right when writing, click on it to close that search.
If you apply any changes (for example, any boosting factor on a product), click on the “Refresh” button for them to appear on the results.
Just below the search bar appears the number of results, listing the products from your data feed; the corresponding query type; and the Relevance Criteria feature, which you can use to alter the natural order of results.
Also, when searching for a product or category of products which are configured with Suggestions, Synonyms, Banners or Custom Results, it will be indicated right under the search bar.
By clicking on top of each configured feature you will be redirected to the corresponding section.
On the main section, you will find four columns: Item, Relevance, Boosting and Score.
-
Item: list of products from your data feed. Click on the plus “+” button to unfold all information from the product. Just beside the plus "+" button, the image, and the name of the product shows up. Click on the name to see the product on your webpage.
-
Relevance: shows the original score of each product according to the search query before applying any boosting factor.
-
Boosting: displays the total boosting value after applying any boosting factor. Boosting is the feature which multiplies the original score (relevance).
To know the value assigned to each boosting (whether automatic boosting, manual boosting, and/or boosting rule) put the cursor on top of the number.
If you don’t see the changes when applying a boost, remember to click on “Refresh”, beside the search bar. Learn how to apply boosting factors by reading about Promotional Tools and Optimize.
-
Score: indicates the total value of a product. The score value determines how close the result is to the search query. The more relevant the product is to the search query, the higher the score. Learn on Score in our glossary or in our FAQs.
The score can be given by the result of Boosting multiplied by Relevance; or, by a Custom Result.
Products with a Custom Result always appear in the first positions as a rocket icon 🚀. Since it is personalized it will always provide the most relevant result.
Query Type Explained
As mentioned above, when testing the outcome of a search, you will encounter a query type. The query type indicates whether the results shown exactly match the search performed, are approximate, or return no results. There are four possible query types: match_and, match_or, fuzzy, and match_all. Keep reading to understand the differences between them.
Match_and
Indicates a perfect correspondence between the search term(s) and the indexed terms. This can occur in both single-term and multi-term queries. In the case of the latter, all searched terms must be present in the indexed content.
Examples:
-
Single term: If you search for "Nike" and receive a query type
match_and, the results will contain an exact match, such as "Nike shoes". -
Multi-term: If you search for “red t-shirt” and receive a query type
match_and, the results will contain an exact match where both terms are indexed, showing only red t-shirts, for instance.
These results are highly relevant to the search performed by the user.
Match_or
Indicates a matching type that broadens the search by returning indexed content that contains at least one of the search terms provided in a multi-term query. This query treats each term individually. As a result, it returns an exact match with the searched terms separately.
Example:
Searching for “red t-shirts” may return occurrences of "red" and "t-shirts" independently, as the exact phrase “red t-shirts” does not exist as an indexed term in the feed.
Match_all
Indicates that Doofinder returns the entire catalog. For example, when using Initial results before a search query is entered; or, when the results returns a Custom Result specifically configured for a search term, since Doofinder doesn't need to analyze the search request.
Fuzzy
Indicates that there is no exact match between the searched term and the indexed term, tipically due to typos or misspellings. For example, searching for "tsirt" instead of "t-shirt". Doofinder will attempt to return the closest possible results, although sometimes it may return none.
Fuzzy terms appear in Zero-result searches within Stats. It is recommended to monitor these terms to apply Synonyms or Custom Results to ensure relevant results. Learn more about Synonyms, and Custom Results.
Unexpected Search Results
When you’re getting unexpected results on the search, it’s important to check and improve your product feed and Search Fields configuration, as both components directly affect the quality of results given by the layer. How can you check that? We give you a few indications below.
Step 1: Verify in Results Preview
Go to: Admin Panel > Search > Results Preview > Search for your specific terms (e.g., “12v”, “50ah”)
Check the Query Type indicator below the search bar:
- If showing
match_ororfuzzy: Your content isn’t matching well enough with what is being searched.
Step 2: Inspect Your Product Feed
Go to: Admin Panel > Configuration > Search Engines > select engine > Indices.
Verify the data:
- Are “12v” and “50ah” present in your product data?
- Which fields contain these terms? (title, description, custom attributes?)
- Are they in separate fields or combined?
- Are those fields included in the Search Fields?
- You can also use the Results Preview to check if your product is showing and the data in it.
Example issues:
- Product title: “Battery” (missing voltage and capacity).
- Specifications are under long descriptions.
- Custom attributes like “voltage” not being indexed.
A quick fix would be to add specifications to product titles: “12v 50ah Battery”, that way you ensure the information searched is actually present on the data of your feed.
Best practices for your data feed:
- Keywords in shorter fields (title, short_descriptions or dedicated custom attributes) score higher.
- Structured data allows better matching.
- Doofinder can more easily find products with ALL terms.
Step 3: Check Field Name Mapping
Go to: Admin Panel > Configuration > Search Engines > select engine > 3 dots next to the indices name > Field Name Mapping
Verify:
- Custom attribute fields are properly mapped.
- Technical specifications are being indexed.
- All relevant fields are included (not excluded).
Example issues:
- Technical specs only in descriptions (it can lose relevance if long descriptions).
- Important attributes excluded from indexation.
Step 4: Search Fields Configuration
Go to: Admin Panel > Search > Advanced Preferences > Search Fields
Search Fields control which fields Doofinder searches in and how important each field is (weight 1-10).
Let’s check again our example “12v 50ah Battery”: If “12v” is in the title (field with less information and using high weight) and “50ah” is in the description (long informational field and low weight), Doofinder might not catch the correct product first.
Verify the data:
- Add technical specification fields to Search Fields, such as “voltage” and add them with appropriate weights.
- Test after changes:
- Go to Results Preview > Search for “12v 50ah” > Check if Query Type improved to match_and.
Step 5: Promotional Tools
If improving the feed isn’t immediately possible use the Promotional Tools to manually control results.
- Custom Results
When users search for a specific term such as “12v 50ah”, your selected products appear first, regardless of natural results.
Learn more: Custom Results
- Boosting Products
Boost products that have all terms, bury products with only some terms, this tool changes the relevance of products.
Learn more: Boosting
- Synonyms
It enhances your results by mapping terms from your feed to different terms of your choice, deriving then to a more accurate result.
Learn more: Synonyms
Always test your changes on the Results Preview, as this is updated immediately after saving configuration!
Note that the Promotional Tools aren’t indicated to resolve a high amount of product codes. In that case, you will need to review your indexed data instead, as informed before.