Relevance
Optimise what product attributes are used across search. Customise the search behaviour by controlling the importance of different attributes like title, type or tags. You can control the configuration of relevance for instant search and search results pages.
What impact do relevance settings have on the site?
The relevance settings will determine the search behaviour and the order in which products appear on instant search and search results pages. It also impacts the products that auto-populate when creating new search curations.
What's the difference between instant search and search page?
Instant search: This usually appears as a dropdown just below the search bar. As shoppers enter a search query into the search bar, the dropdown will feature recommended products. Optimising this is very helpful for shoppers as they are able to see the most relevant products as they type their query in real-time.
Search Page: This refers to the product list page that loads once you hit enter for the search button. This usually takes you to a dedicated landing page that features a list of the most relevant products that relate to the search query.
What determines the products that appear on search pages?
Configuration of relevance
Configuration of synonyms
The way the product data is populated in Shopify
Configuration of search curations (applies to specific search terms)
Setting up the relevance fields
How many fields can I add to the relevance settings?
There is no limit as to the number of fields you can apply but at the minimum, we recommend having all the following fields:
Title
Category
Colour
Make sure all synonym fields are in your relevance settings
This ensures that synonym queries are considered anytime someone enters a term that does not already exist in your Shopify catalogue.
How do I add the synonym field in the relevance settings?
Make sure to check your synonyms, and ensure those synonym fields are in your relevance settings. For most merchants, it means having a synonym field for Title, Colour, Category and Variant Option.
You can now control the relevance level and importance level of every synonym field. Make sure the synonym field is 1 importance less than the main field so that the most relevant products appear first. Here’s an example of a typical setup:
What does “always search” mean?
This means that it will apply to both instant search and search results pages. We recommend selecting “Always search” so that your search output is consistent across the entire user journey.
What importance level should I assign to each field?
Set the importance level between 1 to 10. As long as the high-priority field has an importance level that’s higher than the other field, the actual importance value shouldn’t matter. For example, Title might be 8, whereas SKU might be 2. The higher the importance, the higher the score. Read more about how the score is calculated.
How is the "score" calculated?
The score understands each search match. You can set the title to have an 8 of importance where a certain tag might only have an importance of 2, which means if the search terms were found in a product tag it would score lower then if found in the title, and as such, the product with the match in the title would appear before the other within the results. Score is the sorting from the elasticsearch based on the query, so taking into account boosting and sorting rules, it generates a score and this is used for the ordering, e.g. if you search for "black shoe" and a product with a title "black shoes" it would have a high score, where a product with title "orange and black shoe" would have a lower score.
How does typo tolerance work?
If a search term is only one or two letters off, the relevant results will still appear. That's because your customer's search intent and context are taken into account through typo tolerance, query replacements, and natural language detection. The typo tolerance feature is only configured based on a single word/mis-spelling. e.g. These all return the same:
Black,Blac,BlakDress,Dres,Dresss
What tolerance level should I apply to each field?
You have full flexibility in controlling the tolerance level.
We recommend starting with fuzzy and testing a few search terms to see how it behaves. If you’d like to narrow your results, you can decrease the tolerance level by setting the field to partial. While you can use exact it just means that the results will be a lot more narrow and in doing so, it'll be more accurate.
Why is the typo tolerance not "broad enough"?
Why does Dross not render the same results as Dress when it's obvious that the customer's intent is to search Dress? The answer is that if we configure typo tolerance to be too flexible, the search results can become too broad and inaccurate. Our recommendation is to add synonyms to support multiple word corrections.
How are specific fields referenced by the app?
Reactify Search references your Shopify data to power the following features:
Sort
Filter
Merged Values
Relevance
When configuring any one of these features, you'll notice that the field must be selected. This is so that the app knows which field to reference. Go to Configuration, choose the feature you want to set up, then select the appropriate field from the dropdown list. For example:
Configuration > Relevance > Select a field
The "Field" dropdown will list all the possible Shopify fields you can reference. All the values within that Shopify field will usually appear as a selectable option on the front end of your site.
Collection Title (exact)
Product Type (exact)
Tags Colour (exact)
Tags Category (exact)
Tags Features (exact)
Variant option 1 (exact)
Variant option 2 (exact)
Variant option 3 (exact)
Vendor (exact)
Tags in your dropdown list will appear unique to you, based on how the tags are named in Shopify.
What does Exact, Partial, Fuzzy mean?
Exact, Partial, Fuzzy mean?When selecting a "Field" in the dropdown list, you'll notice another word within the parentheses ():
Exact means perfect match, including capitalisation. So, "ABC" does not match "abc" and if the product title is "ACME Shirt" and you search for "ACME", no results will appear. If you were to select
Title (exact)for your relevance settings, then results will only appear if you enter the exact title.Fuzzy is the most common setting. It means the keyword is analysed. If you select
Title (fuzzy), and your title is "ACME Shirt", then a search for "ACME" will still show results, among other products based on the search algorithm.Partial is similar to Fuzzy, but specifically only a substring, so with
Title (partial), and the term "ACME Shirt", "Shirt" will show up, but "Shirtt" won't show anything.
Partial match, ignoring capitalisation This is usually the default setting and a balance between exact and fuzzy
Exact match, including capitalisation This setting will only serve products that perfectly match the terms so results will be limited
Partial match, ignoring capitalisation and allowing for 1-2 letters different This setting has a higher tolerance for typos, and similar words, thereby rendering more results
“shirt” ✅ “shirt” “shirt” ✅ “Shirt” “shirt” ✅ “t-shirt” “shirt” ❌ “short” “shirtt” ❌ “shirt”
“shirt” ✅ “shirt”
“shirt” ❌ “Shirt”
“shirt” ❌ “t-shirt”
“shirt” ✅ “shirt” “shirt” ✅ “Shirt” “shirt” ✅ “t-shirt” “shirt” ✅ “short” “shirtt” ✅ “shirt”
It's important to select the right tolerance level for your fields to determine the most relevant results across your site. The general of thumb is that Partial and Fuzzy will normally apply to Title and Description. All else will be Exact.
Exact is useful for things like SKUs, where if the SKU is ABC123, then you usually don't want it to come up for a search of ABC, you want the entire SKU code. If the exact term is not required and you're comfortable with having a broader results list, then you can select either Partial or Fuzzy (whichever is the most appropriate).
Enable Search Suggestions
This feature enables you to showcase suggested search terms based on the keywords entered in the on-site search bar. The suggested terms will appear with the most relevant Shopify collection titles or product titles.
Before utilising the suggestions feature, it must be configured by your front-end team. Here are the instructions.
Exclude Search Suggestions
Enhance your suggested search experience with the "Exclude Suggestions" feature, which allows you to omit certain terms from appearing in search suggestions. Note: For the excluded suggestion rules to function, your Reactify Search instance must be on version 5.46.1 or higher.
To exclude suggestions, navigate to Manage > Product > Suggestions
Last updated
