Custom ranking of popular suggestions returned by autocomplete search

đź“– In this article, we will learn more about how to create custom ranking for Boost AI Search & Discovery's Suggestion Dictionary feature.

On this page

Introduction

By enabling the app’s Instant search widget, your search box will be integrated with the autocomplete search drop-down menu. Autocomplete can offer your customers a list of popular searches that they can select from as they type into the search box. 

To adapt our search features to your business, Boost AI Search & Discovery provides a Suggestion Dictionary feature that allows you to customize the ranking of popular suggestions returned by autocomplete search. It helps you streamline autocomplete’s popular suggestions to increase the visibility of some product lines and display them higher in suggestion results.

For example, when customers type “hat”, they will receive a list of suggestions, including "bucket hat", "cowboy hat", and "straw hat". With this configuration, when they search for “hat”, “cowboy hat” will be put at the top of popular suggestions, and the list offered will end with “bucket hat”.

How autocomplete search displays popular suggestions

When your search box is integrated with our Instant search widget. There are two factors influencing the ranking of popular suggestions which are displayed by autocomplete.

  • Priority level: How each record is set at a priority level decides how they will be displayed in suggestion results. There are 3 priority levels that our search features apply to your store data: high, medium, and low. High-priority records are displayed before ones with medium and low priority.
  • Source index
    • Original store data: Popular suggestions can be proposed based on the original data collected from your store. This type of data will automatically clean up after 3 months. The default priority level of records pulled from this data source is medium, with the lowest relevance.
    • Analytics data: Our autocomplete functionality can leverage Analytics data to offer popular suggestions. Our Analytics data can deliver insights on the top 500 most popular search suggestions. These data are cleaned up automatically after 3 months when no longer correct. By default, popular suggestions based on Analytics data have medium priority, with medium relevance.
    • Suggestion dictionary: Once you have configured your Suggestion Dictionary index, autocomplete search will start building suggestions pulled from this source index. Each record that you create in this index will have higher priority than those from the two above source indices. 

Note: Our search functionalities also use other algorithms to optimize the relevance of suggestions returned by autocomplete. 

To give you a better demonstration, every time autocomplete displays a list of popular suggestions, the order of them is as the following:

  1. Records with high priority in Suggestion Dictionary.
  2. Records with medium priority in Suggestion Dictionary.
  3. Records with medium priority from Analytics data.
  4. Records with medium priority from your original store data.
  5. Records with low priority in Suggestion Dictionary.

Customize the ranking of popular suggestions returned by autocomplete search

When you want autocomplete search to display a phrase at the top of the popular suggestion results, you can set it at a high or medium priority level in Suggestion Dictionary.

If you want to hide them from the list of suggestion results, you can set their priority level as low in Suggestion Dictionary.

To customize the ranking of popular suggestions returned by autocomplete search from the app dashboard:

Note: The following tutorial requires the Instant search widget integration into your search box with the Popular suggestions component enabled.

1. From the app’s admin, go to Search > Standard Search > Select Suggestion dictionary from left sidebar.

2. Click + Add new. 
3. Enter a suggestion word or phrase that you want to customize its ranking in the suggestion results in the Term box of the Suggestion column. Select its priority level in the corresponding drop-down of Priority. You can add limitless suggestion terms. 
In the example of "hat" above, if you want "cowboy hat" at the top and "bucket hat" at the bottom of popular suggestion results return by the autocomplete search box, you can implement the feature as the following:

4. Simple enable/disable the suggestion terms then Reload your store to check if they are applied to your search box.
5. To manage your Suggestion dictionary:
  • On the Suggestion dictionary page, toggle the button of any suggestion terms to off if you no longer want rules of Suggestion Dictionary to apply to them.
  • Click Edit on the Suggestion dictionary page to either customize the term and its priority level or delete it completely. Click Save suggestions after finishing the customization.
Note: There are 2 types of suggestion terms you can consider adding to Suggestion Dictionary. The first is the product categories from Analytics data that you want to push to the top of suggestion results, the second is search terms you think customers will search for the most in your store.

Import and export your Suggestion Dictionary

Our app allows you to export a large number of suggestion terms at one time. This feature can help you to create a backup of your Suggestion Dictionary, exchange suggestion terms between different Shopify stores that have Product Filter & Searched installed, and edit them in bulk.

To export the Suggestion Dictionary from your Shopify store:

  1. From the app’s admin, go to Search > Suggestion dictionary.
  2. Click Export under the Suggestion dictionary heading.
  3. Click Export in the Export suggestion dictionary to CSV file pop-up to receive the CSV file of your existing suggestion terms.
  4. Double-check the file that was downloaded by your browser.
  • Note: When your Suggestion Dictionary has more than 50 suggestion terms, the CSV file will be emailed to your store's owner.

After exporting your Suggestion dictionary and making changes to the CSV file, you can import the updated CSV file into the app's admin. To import your Suggestion dictionary using a CSV file:

  1. From the app’s admin, go to Search > Suggestion dictionary.
  2. Prepare your CSV file with suggestion terms that you want to import into the app. You can download the sample to understand all fields that your CSV file can contain. 

You can use Notepad, TextEdit, or any similar text editing app to check if the format is correct.

  1. Click Import under the Suggestion dictionary heading.
  2. Click the Choose file button and select the CSV file that you want to import. 
  3. Click Upload.
  4. You will receive a message about new suggestion terms that the CSV file contains. Check this notification to make sure all suggestion terms that you want to import are recognized. Then click Import.
  5. After successfully importing, you will receive the following message:

  1. You can continue to import other suggestion terms by clicking Import more suggestions and repeating the above steps.
  2. Select View all suggestions to make sure all the suggestion terms have been added to the app’s admin successfully.


If you have any questions or need further assistance, please do not hesitate to reach out to our dedicated support team at support@boostcommerce.net.