Search and Navigation Implementation Considerations
Guidelines and requirements you may want to consider as you implement search and navigation throughout your storefront.
Identify how many types of search are required for the solution. For example, do you want different behavior or presentation for different categories: top sellers versus clearance items or for content assets versus products?
- Determine which pages show search results.
- Determine how many search results to show per page.
- Determine what happens when:
- No search results are defined for a category or product
- Less than the planned number of search results are defined for a category or product. For example, you want to redirect a search to a category page or another page to provide the user with more results.
- More than the planned number of search results are defined for a category or product. For example, you want the customer to be able to refine their search.
Sorting and ranking logic:
- Determine how many ways a customer might want to sort results
- Determine whether you want to promote specific products through search
- Determine what happens when:
- Multiple sorts are applied in a specific order
- Ranking logic is applied to reposition results after sorting
- Identify what additional information is required for indexing, so that customers can search on it. Every attribute that is needed for indexing must be marked as searchable.
- Identify required search features, including:
- Category placement
- Search placement
- Search rank
- Text relevance boosting
- Identify how data for indexes will be maintained (for example, import, manual, programmatic, or other):
- During development
- When the site is live
- If the merchant plans to manually create search indexes in the Business Manager, identify:
- Who will be responsible for creating them
- If there will be a review process
- How they are tested
- How search indexes are replicated with catalog data
If the merchant plans to import search index definitions, define the import process. If possible, implement it as a job within Salesforce B2C Commerce, so that you can take advantage of the scheduling and notification functionality built into B2C Commerce. If you intend to perform regular index updates, these can also be scheduled in Business Manager.
During the design phase, merchants should consider the day-to-day operational overhead associated with creating and maintaining search indexes manually within B2C Commerce, especially merchants with large product catalogs. This process includes examining current customer searches, identifying needed changes, configuring dictionaries and indexes, and manually building and replicating indexes.
Build code to support search requirements.
To facilitate development and testing, you must:
- Create users with permissions that are in line with merchant-defined business processes.
- Create an initial set of test search results.
- Use a test plan to ensure that all use cases are reviewed, including current Storefront searches, if known.
- Test the import process, if applicable.
Salesforce recommends the following best practices:
- Show all variations for sale; don't attempt to narrow the results by variation using scripts to process search results. For example, if color is a slicing attribute, show all iPods in all colors, don't attempt to iterate over the results to pull out only pink iPods via scripts. While this can be done, it requires post-processing of the full result set, which can impact the search performance. Instead, it's recommended that you add a search refinement that lets the customer drill down a specific selection.
- Any result processing that requires that you iterate through all the results can make search a lot slower.
- Be careful when applying text relevance boosting. Because this feature is site wide, it can have unexpected consequences if not applied with the full search set in mind.
- Make sure that scheduled indexing is disabled when you replicate index information from staging to production.
If you have a large catalog of products and you want to improve the speed of search, try to make the size of your indexes smaller. You can see the size of an index after it's built on the Search > Search Indexes page. Usually, the number of attributes being indexed is what determines index size. You can also add stop words to decrease the size of the index.
However, if you decrease the number of searchable attributes, this means that the search results might be less accurate.
If you have an average size catalog of products, decreasing index size isn't normally a concern.
We recommend testing the top 50–100 searches on your site as a baseline to make sure that the search results are returned in the order that you prefer.
Consider the following limitations when designing search features.
- Parallel product searches and contextual searches are currently not supported.
- Salesforce B2C Commerce currently supports only one index based on real-time data: the availability index. This index, based on an inventory list can support one or more sites. Multiple sites that use the same inventory list also use the same search availability index.
You can use the productgrid.isml templates in the SiteGenesis application to control the number of search results returned and the display of search results. You can use the productnav.isml templates to control the look of the paging bar that has the previous and next buttons, and the links to each page of results. Both templates are located in the SiteGenesis Storefront Core cartridge in default\search. The SiteGenesis application includes customer paging preferences.
We recommend you test the quality of your searching regularly and we’ve prepared some resources to help with your testing.
You can use these resources with the Reports & Dashboards Search Conversion reports.
- Quality of Search Feedback Form for Reports & Dashboards (after July 13, 2020): A Quality of Search Feedback Form to be used along with the Best Practices of Site Search Optimization methodology.
- Quality of Search Feedback Form for Business Manager Reporting (before July 13, 2020): A Quality of Search Feedback Form to be used along with the Best Practices of Site Search Optimization methodology.
(Formally Business Manager Analytics)
These reports are being deprecated. You can use the Historical Reports module to generate reports and review data for dates before January 1, 2017.
- Quality of Search Template How-To (1 of 2): This video walks through how to download the Search Conversion Report from Business Manager and import it into the Quality of Search Testing Tool Template.
- Quality of Search Template How-To (2 of 2): This video walks through the Quality of Search Testing Methodology and how to execute it using Business Manager tools and reports.