Product Recommendations FAQ

General

What do I need to do before I can use recommendations?

At the minimum you need a product catalog that adheres to the new product catalog specification. To maximize functionality, you should also update your Monetate implementation so that it collects SKU-level data on a page.

Can I make recommendations based on stock level?

Although stock level, frequently called quantity, isn't part of the base product catalog specification, Monetate allows custom attributes to be loaded in the feed. Therefore, if you include a quantity custom field when creating your feed, you can send through data about how much product you have available to sell and leverage that in your filters.

Will there be results in a recommendation strategy immediately after creation?

Recommendation strategies that use algorithms to generate results may require overnight processing to generate results. You can check the status tab from the Recommendation Strategies list page to see whether results have been generated.

Can I use recommendation strategies from multiple catalogs?

Ensure that you are using strategies from a single catalog. Using strategies from multiple catalogs can produce unexpected results.

Why am I seeing these products in recommendations?

Monetate generates a dataset based on item group ID (PID) and expands that to a group of IDs (SKUs) for filtering before showing them as the recommendations.

If multiple recommendations are running on a page, does Monetate prevent duplicate products from showing across each recommendation?

Monetate uses the priority ranking of the actions and experiences to deduplicate. Products displayed in the highest-priority action are removed from lower-priority recommendations running on the same page.

However, if a client has a hybrid implementation and is running both an Omnichannel recommendations experience and a Monetate tag-based recommendations experience on the same page, the deduplication occurs within each respective request and not across both of them.

Recommendations and Product Catalogs

How should I configure pricing for sale items in a product catalog?

The value of sale_price should be the lower price. You should still include price in the product catalog.

Do I need to collect SKUs on a page to use recommendations?

No. You can still base recommendations on PID-level data, but to unlock the full capabilities of the product, you should collect and use SKU-level data.

How do I update a recommendations product catalog via SFTP?

You can automatically update your product catalog via SFTP. Refer to Update a Product Catalog via SFTP for detailed instructions.

If I have an error in one required field, does the entire recommendations product catalog fail?

Yes. If there is an error in any required field, then the entire product catalog fails validation. No data will be populated from the new catalog file.

What happens if a recommendations product catalog fails to upload?

Monetate retains the data from your last successful upload. Be aware that an error in a single row will cause the entire catalog file to fail verification and will result in the upload being discarded.

Do I need to pass SKU data to leverage dynamic filtering?

No, but doing so makes the feature more powerful. You can still use PID-level data to use dynamic filtering.

What IDs does Monetate deduplicate on?

The values from the item_group_id field from the feed should match the PIDs sent via the API method on the webpage. They are parent group IDs. The ID values sent in the feed should match the SKU sent across via the API method on the webpage. SKUs are the children of the PID. Monetate deduplicates by item_group_ids.