Merchandising

Merchandising is a way for you to affect the results of searches and recommendations by boosting, burying, filtering or pinning products, product categories, content pages and content categories. This can help you advocate certain products that you know have a high conversion rate but low viewership, or bury some products that are prone to getting returned and refunded.

Merchandising is a powerful tool that lets you improve the visibility of important products and pages, or filter out products you do not want to show in your search results. Learning how to use Merchandising rules enables you to get the most out of your Relewise experience, and lets you fine-tune your search and recommendation algorithms to perfectly fit your needs.

To learn more about the Merchandising interface, click here.

Scheduling

Scheduling lets you automate when a rule is active, giving you full control over its timing. Set a specific start and/or end time to ensure your merchandising strategy runs exactly when needed.

Rules without a start time take effect immediately, while those without an end time remain active continuously. This flexibility allows you to plan ahead, knowing your rules will activate and deactivate at the right time.

Typical use case

Scheduling is useful for time-sensitive rules, like boosting sale items during Black Friday or promoting seasonal products.

Boost/Bury

A boost/bury rule affects the relative weighting of an entity, such as a product, in a positive or negative direction, thus causing the entity to either become more or less prominent in the search or recommendation results they might be shown in.

Specifically, boosting and burying is done on the basis of the entity's relevancy score, which is calculated off of its popularity between sales and views, and the direct relevance to the user viewing it. This means that a boost or bury score is relative to the score that exists on the entity, and should not be considered as a fixed value.

For instance, if product A has a relevancy score of 50, and product B has a relevancy score of 10, a boost of 10% will have a disproportionate effect on the two: Product A's relevancy becomes 55, while Product B's relevancy becomes 11.

To learn more about boosting and burying, click here.

Typical use case, Boost

Boosting is useful when a certain product or brand has a high profit margin, and you want to promote it in searches and recommendations.

To learn how to boost on product profit margin, click here.

Typical use case, Bury

Burying is useful when a product is getting returned a lot, and you want to lower its ranking compared to other, more stable sellers.

Boosting/Burying Variants

A use case may arise where you want to boost a specific variant of a product in either a search or recommendation scenario. Depending on which, the method differs. Note that the following logic applies to both boosting and burying variants.

Boosting Variants in Term-Based Searches

It is possible to boost a variant directly in a term-based search by targeting its variant ID. Doing so will cause that specific variant of the product to rise towards the top of the search, while leaving any other variants of the same product untouched.

Boosting Variants in Termless Searches

Since Termless searches will always display variants of the same product one after the other in the search results (the PLP), boosting a variant will invariably boost the product as well, and implicitly all other variants of the product. This is, naturally, contingent on having set a ExplodedVariants value greater than zero.

Boosting Variants in Recommendations

Due to the way recommendations evaluate products before variants, it is not possible to boost a variant in isolation. Basically, a boost rule set up to target a variant will only have an effect if the products associated with the targeted variants have already been selected to populate the recommendation results.

For this reason, in order to ensure that a variant is boosted correctly, it is also necessary to boost its associated product.

Filter

A Filter rule prevents entities from being shown in searches and recommendations. Please note that the Filter Rule works as an exclusive filter, meaning that it filters all the selected entities out of search and recommendation results. This is the opposite of Relewise's general request filters, which are inclusive and define which entities are allowed to be returned in results.

Typical use case

Filter rules are typically used to filter out products that are out of stock.

To learn how to filter out out-of-stock products from request results, click here.

Pin

The Pin rule is used to lock a particular entity to a specific position in a recommendation slider. Please be aware that the Pin rule only affects recommendations at this time.

A pin rule cannot be used to cycle a specific position between a set list of products. To learn more, click here.

Typical use case

When you want to display your best-selling product on the power step to boost visibility.