Personalization Settings include both Exclusion Rules and Advanced Rules.
Exclusion rules empower merchandisers to exclude specific products from Personalized Recommendations. There are three types of exclusion rules:
- Global Exclusions are used to remove products from all Personalized Recommendations across a site.
- Contextual Exclusions exclude products from being recommended in a specific context.
- Contextual Groupings enable merchandisers to group products together so that they are only recommended when shoppers are viewing a product that is a member of that group.
Changes to these settings will only take effect after the Personalization Generator runs an update. These updates are scheduled to occur at least once per day, commencing immediately after your product data is updated.
Global Exclusions
To create a global exclusion, choose an attribute from the customer's product data and the value within the field that you wish to exclude on, and products containing that value will be completely excluded from recommendations.
Contextual Exclusions and Contextual Groupings
To create a contextual exclusion, choose an attribute from the customer's product data and the value within the field that you wish to exclude on, and products that contain that value will only be recommended when the shopper is viewing products that also contain the excluded value.
Contextual groupings will dynamically update as product data changes to account for new values. The result is the same as the contextual exclusions described above. The difference is that contextual exclusions create one group for each rule. Contextual groupings can create many groups for each rule, since a separate group will be created for each unique value of the selected field.
Products that are not covered by a contextual exclusion or contextual grouping are considered “universal” and can be recommended in any context.
Exclusion Rules Use "or" Logic
Contextual exclusions and contextual groupings work independently; they do not naturally unite. As an example, we can exclude sale products or a brand grouping from recommendations. However, an on-sale exclusion and a brand grouping won't guarantee that recommendations will show only full price products from the same brand.
We advise against creating a contextual exclusion and contextual grouping for the same group of products. If you do set up exclusion rules in this way, they may conflict rather than compliment each other, and they can lead to unfavorable results.
What Exclusion Rule Should I Create?
Here are some common use cases.
| Objective | Exclusion Rule | Examples & Outcomes |
| Do not recommend sale items with full-priced items | Contextual Exclusion | on_sale is yes: Only recommend sale items when looking at other sale items |
| Do not recommend out-of-stock products | Global Exclusion | In Stock (availability) is Out of Stock: Do not recommend out-of-stock items across the site |
| Do not recommend products with a price lower than 0.01 | Global Exclusion | Price (calculated_price) is less than 0.01 |
| Only recommend products from the same brand as the seed product | Contextual Groupings | Brand: Only recommend products of a given brand when viewing other products from that brand |
|
Warning: Searchspring strongly discourages global sale exclusions that are configured as in the example below. It is possible to break your recommendations functionality with a site-wide sale because this rule would exclude all products.
We recommend creating a contextual grouping for sale items instead, as shown below. This rule will restrict sale items so that they are only recommended when the shopper is viewing another sale item.
|
Generator Status
The Generator Status page shows the status of recommendations updates. Typically, recommendations are updated after your product data is indexed. If there has been an error with updating recommendations, those errors will be reflected here. This page also allows you to manually re-generate product recommendations if you have made a change to your configuration settings that needs to be processed before your next scheduled update.
Upsell Filter
The Upsell Filter applies a price filter to Similar recommendations so that all profiles using the Similar recommendations type will recommend products at a higher price than the seed product’s price. This filter will apply to all Similar profiles after it is set; it cannot be set on a profile-by-profile basis.
Exclusion Statistics
Once you set up your Exclusion Rules and the Personalization Generator runs successfully, this page should now have a space for Exclusion Statistics. Its objective is to report on the health of your Exclusion Rules.
Exclusion Statistics report the number of products that are being excluded per each Exclusion Rule. It also outlines the total number (and percentage) of all excluded products, and it will warn with caution if this percentage is higher than the recommended threshold.
Advanced Rules
Merchandisers have the ability to import and export advanced personalization product rules, brand rules, and category rules. The import and export options are found in the Searchspring Management Console under Personalization / Personalization Settings / Advanced Rules.
Note: This function may be especially helpful for merchants new to Searchspring who want to import their existing complete-the-look rules.
Import button
Click the Import button (which displays an up arrow) to import rules. You will see a message during bulk import asking for the product ID, which is typically the product SKU, but the product ID will also work.
Note: Importing new rules will replace all existing rules in the selected section.
Export button
Click the Export button (which displays a down arrow) to export rules.
Note: In the image shown below, because there are no brand rules or cross-sell product rules configured, those sections lack an Export button, which will appear only if there are rules to export. The Import button will always display.
Category Rules
Category Rules can be used to curate the type of products that will be recommended in cross-sell or cart cross-sell profiles. Simply select a seed category and a featured category. When shoppers view products that are members of the seed category, then we will recommend a product from the featured category.
Merchandisers can set up multiple category rules for a cross-sell profile. As an example, let's say you use a seed category of "pants". For the "pants" seed, you can create a featured category for "shirts" and another featured category for "belts". Now when shoppers view pant products, we will recommend a shirt as the first cross-sell result and a belt as the second cross-sell result.
Category rules will boost products to the front of a cross-sell or cart cross-sell recommendations profile, taking precedence over product pins.
You may also block a category from being recommended when shoppers view a product associated with the seed category.
As an example, let's say a merchandiser wants to encourage shoppers to purchase entire looks. More specifically, when shoppers look at pants, the merchandiser wants to recommend shirts. To do this, the merchandiser creates a category rule with a seed category of "pants" and a featured category of "shirts". That rule will go into effect once incorporated into the recommendations generator.
NOTE: If there are no recommendable products from the featured category, it will be ignored and the system will default to recommending products based on the individual shopper's history and aggregate crowd data.
Brand Rules
Brand Rules can be used to curate the type of products that will be recommended in cross-sell, cart cross-sell and similar recommendation profiles. Similar to Category Rules, you can create a seed brand and a featured brand. When shoppers view products that are members of the seed brand, then we will recommend products from a featured brand. You may also block a brand from being recommended when shoppers view a product associated with the seed brand.
When a shopper is viewing a Nike product, you may want to block recommendations of Adidas products. To do this, you would create a brand rule on the brand “Nike” to add “Adidas” as a blocked brand. After that rule is incorporated into the recommendations generator, Adidas products will not be recommended in cross-sell or similar profiles when shoppers are looking at Nike products.
Product Rules
Product Rules can be used to boost or block specific products in Cross-Sell or Similar profiles. Based on a specified seed product, you can either boost a specific product to ensure it’s recommended alongside that seed product or block a specific product to prevent it from being recommended for the seed product.
You can create separate product rules for cross-sell recommendations and similar recommendations, as promoted cross-sell products may not make sense as similar products.
Category Rules, Brand Rules and Cross-Sell Product Rules apply to both Cross-Sell and Cart Cross-Sell recommendation types. Any profile using either of these types will be affected by rules set here.
Comments
0 comments
Article is closed for comments.