Implementing Regions

Overview

Implementing the use of regions within Algonomy allows merchants to pass different product attributes, such as list price, sale price, catalog names and availability, for those regions. Regions will also allow merchants to use different currencies in those different locations.

Using regions allows users to use regional strategies, set different currencies and even different languages, based on the instrumentation on the page. Support for regions requires the use of a Regional Feed as well as the instrumentation of regions on all pages.

When regions have been implemented, the recommendations that are returned are adjusted to reflect the regional values. For example, if regions have been implemented and a user in one region is viewing recommendations, those recs will reflect the price, availability and currency of the region he/she is shopping from.

There are a number of region-aware strategies that can be enabled to take advantage of the region implementation. For all other strategies, regional data is substituted after products are chosen (the strategy has considered customer behavior for all regions on the site). Sites that use regions will return effective recommendations only if there is a large overlap of products between regions.

If a site offers many products that are specific to a single region, products from the larger regions will dominate the recommendations and the system will not be able to recommend products that are available in smaller regions.

Regions are also required for localization. To support multiple languages and currencies, regions must be implemented.

Setup

To add region support, you need to:

Tell your Algonomy team so that regions can be enabled for your site and set your default region.
Update the feeds. The variable values are sent through the regional feed and associated to a region ID. See Product Region Flat File Feed for a complete list of attributes that you can set for a region.
Update your site integration. In the integration logic, you identify the user's region ID. When the region ID is known for a user, the recommendations that are returned use the available values for that region.

Notes:

The customer does not need to be logged in to be assigned to a region.
Region ID can vary during a session.
Region IDs are case sensitive.

Choosing the Default Region

Sometimes, there's no region-specific information to show a customer. This can happen for different reasons:

The feed didn't contain details for a product in a particular region.
A valid region ID wasn't passed with the instrumentation or API call.

When there's no regional information to show, recommendations show the default region, set up by your Algonomy team.

Recommendable Products

Flat Feeds

For products to be recommended by our system they must have the recommendable flag set to true within the product full file.

If regions are enabled for the site, then an entry for that product must exist in both the product full file and supplemental region file and must both be set to recommendable / in_stock true for the product to be recommended.

Here are some examples:

Product Full File recommendable field	Region File in_stock field	Can it be recommended?
true	true	yes
true	false	no
product not listed	true	no
true	not listed	yes

Regions and Recommendations

In general, region data does not change the recommendations that are returned (except for products that are out of stock in certain regions). Regional information is substituted after recommendations are generated for a user. Backfill strategies become especially important when you're working with sites with multiple regions.

Regional Product Attributes

These values are substituted when a region has been identified:

List price
Sale price
Availability
Margin

Region-Aware Strategies

Clients who have regions instrumented can make use of the region-aware strategies to see recommendations based on behaviors captured within the region. For more information, refer to Strategies Library: Region-Aware Strategies.