Manage Boosting Rules

This section includes instructions for managing rules when using product boosting.

To manage boosting rules:

  1. On the Omnichannel Personalization dashboard, go to Recommendations> Recommendation Boosting. The Recommendation Boosting page is displayed with the rules already created.

  2. By default, all rules that have expired are hidden. Select the Show Expired checkbox to view them.

  3. Click in the Production column to enable the rule in the production system.

  4. Click in the QA/Integration column to enable the rule in the QA/Integration system.

    5. The QA/Integration is Algonomy environment that is parallel to the production environment. You can enable your production website to retrieve recommendations from our QA/Integration environment by adding the following query string parameter to your website URL "r3_env=integration" For example, if your website URL is

    http://www.mystore.com/product?id=12345

    then it will look like this:

    http://www.mystore.com/product?id=12345&r3_env=integration

  5. To learn more about query strings, visit: http://debugmode.net/2010/05/09/simple-explanation-for-very-beginners-of-query-string/. This method allows you to verify the rules behave the way you expected them to, before enabling them for all your customers. If you have any questions, contact Algonomy Support team.

  6. Select the checkboxes corresponding to the rules you want to delete.

  7. Click Delete Selected.

Boost Rule Behavior

When a product is eligible for a boost, that product has its score in a given model multiplied by a static co-efficient and is then re-ranked. This usually puts the product at the top of the ranking list, but lower ranked products may not always end up at the top after a boost.

What happens if you have two products in a model that have a boost rule applied to them?

  • When two products have a boost applied, those two products get re-ranked individually.
    • For example, Product A is originally ranked 5^th^, while Product B is originally ranked 6^th^. Both products are boosted. Product A is now 1, Product B is now 2.

What happens if you have a product that has multiple boosts applied to it?

  • When a product has multiple boosts applied to it, it gets an even bigger boost than if only one was applied.
    • Using the example from above, if Product B gets another boost applied, it will be re-ranked again and become the top product, with Product A falling to 2.