Search Suggestions

When a shopper starts to enter a search term on the storefront, the storefront application shows terms that can help complete the search. Depending on the configuration, search suggestions can range from being a simple list of related terms to being more complex and providing extra details.

The ‘Search-As-You-Type’ feature isn’t controlled by a preference.

To enable Search Autocorrections: navigate to site > Merchant Tools > Search > Search Preferences and select Search Autocorrections.

When enabled, the spelling of the words in the user's search phrase are automatically corrected or completed before the actual search is executed. When disabled, the terms in the user's search phrase aren’t corrected or completed.

You can obtain a corrected and completed version of the phrase using the Did You Mean option.

Set the minimum number of characters for the suggestion length at, site > Merchant Tools > Search > Search Indexes > Language Options.

Search Suggestion Types

If the shopper selects a category or subcategory in the top or left pane of the storefront, the returned suggestions, spelling corrections, and completed terms are all specific to selected category or subcategory. Search filtering by category is available for product, brand, and category suggestions. Content and suggest phrases aren’t category-specific.

Your storefront application can suggest the following types of content:

Additional phrases
Categories
Brand names
Products
Content

You can list suggestion types by group and link to specific landing pages. For example, when the shopper clicks a suggested category, the respective category landing page opens. When the shopper clicks a suggested phrase, a text search is performed using that phrase. When the shopper clicks a brand name, the brand landing page opens.

You can show category, product, and content suggestions with additional information, such as an images, a product category, or a price.

The shopper's search phrase is matched anywhere in the searchable content. For example, when a shopper enters shoe into the search box, a product red sport shoes is suggested.

Linguistic search rules such as synonyms and stopwords are also considered when searching for relevant suggestions. Search phrases are automatically corrected and completed, resulting in suggestions that are more relevant.

Commerce Cloud Einstein Search Recommendations

Commerce Cloud Einstein can improve searches with personalized type-ahead guidance. When enabled, this feature provides AI-driven autocorrection and term completion.

For information about Einstein Search Recommendations and how to enable it, see Commerce Cloud Einstein Search Recommendations.

Note: None of the Einstein search recommendations are affected by the suggestion term blocklist, and they are not guaranteed to yield results.

Simple Search Suggestions

For simple search term suggestions, the search field provides a list of up to 10 terms that help complete the query. The list of provided terms updates with each letter the shopper types into the search field. The search query uses any suggestion the shopper selects.

Search Suggestions Terms

Search entries and suggestions are case insensitive. That is, entering a search term in either upper or lower case doesn’t affect what search suggestions appear. In addition, search suggestions always appear in lower case to avoid duplicates based on capitalization.

When creating suggestions, keep the following in mind:

If both singular and plural values are added when creating suggestions, the singular value is treated as a duplicate and only the plural form of the suggestion appears. For example, even if you create blouses and blouse as search suggestions, only blouses appears as a suggestion.
If the plural form of the word has a different stem, both words appear as suggestions. For example, scarf and scarves.
If you add only the singular form of a word as a suggestion, only that form appears as a suggestion.

All characters are retained from category names for search suggestions, including hyphens and ampersands. The full name of categories and brands always appears in the search suggestion.

Salesforce B2C Commerce automatically includes the following in the Suggest index:

All brand names for all products in the storefront catalog
All display names of all categories from the site's storefront catalog

You can also add terms or prevent terms from being suggested using the Search Suggestion feature in Business Manager. For example, you can add fringe as a suggestion for a new line of clothing or prevent the term Adult Videos from being suggested to shoppers by adding it to the blocklist.

Note: The suggestion term blocklist affects only suggested terms.

German umlaut and eszett characters (ü, ß) and standard replacement characters (ue, ss) are supported. Search suggestions appear as they’re stored in the product catalog. For example, whether you enter träger or traeger, you see trägertops and träger-tops as suggestions for both.

Hit Count

Beside each search term, the shopper sees a hit count. The hit count indicates the number of products or articles that search has found in the storefront. Suggestions are sorted by hit count.

Note: The hit count doesn't consider the availability of the product, so it's possible for customers to click suggestions for products that are currently out of stock.

If you take products that are unavailable offline and rebuild the Suggest Index, these hits disappear. You can also customize the no results page to indicate that the selected product is out of stock and suggest alternatives. If you don't want to show the hit count, a developer can alter the ISML template to remove the hit count.

Search Suggestion Limitations

To help improve performance, the suggestion index includes these term limitations.

Suggestions that don't provide results aren’t included.
Query results with more than 50 characters aren’t included.
When adding a custom search suggestion, the term must appear in indexed catalog data. Providing search suggestions for indexed content data is not currently supported.
Term completion and auto correction is skipped if the search phrase includes wild card phrases, negative phrases, or quoted phrases.

Differences Between Simple and Enhanced Search

This table calls out differences between simple and enhanced search term completion.

Term Completion Feature	Simple Search	Enhanced Search
Category, brand, and extra phrase support
Product and content page support
`Search as you type` support for products, content, categories, and brands
Match support	Whole phrase match For example, suggests `digital camera` when shopper types `digit`, but not when shopper types `camera`.	Independent word match For example, suggests `digital camera` when the shopper enters `camera`.
Hit count provided Note: Hit counts are precalculated and can be outdated.
Autocorrection and autocompletion support Note: For example, both terms of the input text, `digit cam` auto-complete to`digital camera`.
All standard features used for storefront search are used for enhanced search suggest, such as compound words, stop words, and sorting rules.