The way an eCommerce site displays products for a query or which products get displayed relies highly on the relevancy setup. Relevancy settings is the first thing you should visit post integration in order to define the behaviour of your search.

Every search is handled using the following two steps:

Finding the relevant products for the typed query
Ranking the proper product attributes to display the top ones on the page

Relevancy settings comprises defining searchable attributes, providing exact match keywords, adding attribute filters, mapping the catalog fields to Unbxd fields, and detecting user intent.

Searchable Fields

The schema file is full of product attributes, so it is important to index the ones that will affect the search results because all of them will not. So, configuring the fields as either ‘searchable’ or ‘non-searchable’ is important.

Searchable field: When you define a product attribute (like color) as a searchable field, it affects the search results when a shopper searches for a product.

Non-Searchable field: If an attribute is set as ‘Non-Searchable’, then it doesn’t affect the search results, therefore, simplifies the search results by showing only the relevant ones.

For example, the field “description” is marked as Searchable (High, Medium, Low). Now, when a shopper wants to search for a white shirt, he gets numerous other options which are unrelated to the shopper’s query. Like, he receives ‘Jeans’ in the search results because its description had: “can be paired well with a white shirt”. Unknowingly this search result will also be displayed even though the shopper is not looking for jeans. Such instances make the search journey harder for a shopper. Therefore, add them as ‘Non-Searchable’.

Examples of which attributes you should make searchable and which to exclude are:

To include: name, brand, features, color, size, keywords, tags, SKU, ISBN etc.
To exclude: description, for information only (not yet available, not yet released), or ranking status (most viewed, most used etc).

Assigning Search Weights

The way you rank the product attributes on the searchability quotient determines the effect on search results. This means when the engine runs to show up a product based on its attribute the results highly depend on ordering. The higher the rank the more is the participation in ordering the products.

For example, If the query says ‘blue’ and the priority given to the attribute ‘color’ is high and ‘name’ is medium in the relevancy ranking. Now, during a search, if a record X matches with the color and record Y matches with the name, then record X will be placed above in the results page due to the priority ranking.

You can view the relevancy ranking on this screen. Relevancy Ranking screen contains the name of the Attribute and its respective Search Weight.

Attribute: Each product can have a number of attributes for it. Like, color, brand, size, etc.

Weight: The measure of importance assigned to the attribute. You can choose from: HIGH, MEDIUM, LOW.

To view Relevancy Ranking, navigate through Manage > Search > Searchable Fields.

Weightage	Description
Non-Searchable	Attributes that shouldn’t affect the search results should be tagged as ‘Non-Searchable’
High	When an attribute is marked as a top priority to affect the search results
Medium	When an attribute is marked as a second-top priority to affect the search results
Low	When an attribute is marked as the lowest priority of all to affect the search results

Configure Searchable Fields

You can edit the existing weightage of any of the product attributes to give it a higher priority or a lower one depending on your sales needs. To edit the existing weightage follow these steps:

Navigate to Manage > Search > Searchable Fields.
To edit the search weightage, select the search weight from the column against each attribute. You can choose from Non-Searchable, High, Medium, or Low
Click Save Changes.

You have successfully edited the search weightage for an existing attribute.

The details of the AI recommendations are:

Attributes	Description
Product Coverage	Specifies the percentage of how many products are covered under the searchable field.
Query Coverage	Query coverage measures the overlap between search query used by shoppers and information present in the field.
AI recommendation	This displays if the AI engine recommends this field as a highly searchable field, normal, or a low searchable field

If you upload a new catalog or add some new fields, then click ‘Refresh AI Recommendations’. This will update the recommendations based on the new query data. Once done, the screen will display ‘How many fields were added or updated’.

Did this answer your question?
Yes!
Nice work, I love it.
Feedback is submitted

On this Section

To create a Query Rule
To Edit a Query Rule
Delete a Query Rule
Campaign States
Create Campaigns
Edit Campaigns
Preview Campaigns
Duplicate Campaigns
Delete Campaigns

Relevancy Settings New

Searchable Fields

Assigning Search Weights

Configure Searchable Fields

Yes!

On this Section