Recommend products (Legacy)

  • Updated

Personalize product recommendations to encourage sales, improve conversion rates, and increase order value. In Optimizely Data Platform (ODP), you can dynamically generate entire sections of emails or web content to make campaigns more relevant to customers.

ODP uses the behavior of each customer to determine dynamic, personalized recommendations. This feature is ideal for:

  • Campaigns for long-time customers with a history of interaction. 
  • Relatively stable product catalogs and consistently visited or purchased products. 

This feature is not ideal for:

  • Campaigns for customers without much activity, such as new customers.
  • Campaigns for customers with outdated activity.
  • Highly-diverse product catalogs where many products receive little interaction or sales and only a few attract much of the activity. 
  • Seasonal product catalogs where recommendations based on past customer interactions might not be applicable.

ODP has the following default production recommendation options:

  • Optimizely Recommendations
  • Best Selling Products
  • Most Viewed
  • Revenue Generators

Optimizely Recommendations creates automated recommendations based on each customer's behavior compared to similar customers by using collaborative filtering and matrix factorization to predict a customer's response to another product. The machine-learning algorithm runs nightly using event data from the previous 180 days then recommends products with the highest predicted response.

If a customer does not have enough behavioral data to make a prediction, ODP uses default recommendations instead, which are defined by the best selling products from the last week.

As ODP collects additional behavioral data, the default recommendations are replaced with more specific ones.

If you understand your customers or want to control your customer's product journey more, consider using Best Selling Products, Most Viewed, or Revenue Generators. You can also create custom recommendations by working with your customer success manager.

The following recommendations are not personalized for customers but are useful in specific nudges or win-back campaigns.

  • Best Selling Products uses the products that customers purchase the most to make recommendations. 
  • Most Viewed (email only) uses the products that customers view the most to make recommendations.
  • Revenue Generators (email only) uses the products that currently generate the most revenue for the business. 

If the default recommendations do not fit your situation, you can also use product behaviors.

Add to web content

On-Site Recommendations is an optional platform add-on. If you are interested in the functionality, contact your customer success manager.

Use the following steps to add a section that is based on recommendations to web content:

  1. Go to Campaigns. 
  2. Click the name of the campaign you want to edit or create a new campaign.
  3. Select the Edit icon of the web touchpoint you want to personalize
  4. Drag a Product Grid into your content from Elements > Layouts.
  5. Select the default recommendation from Source. If the Best Selling Products recommendation is selected, define the desired timeframe. If the Optimizely Recommendation is selected, choose the desired reference point explained below:
    • Products On Page – Highlights products similar to those on the current page. Use this option to suggest other related products that the customer might enjoy.
    • Recent Browse History – Highlights products that the customer viewed in the last 2 weeks. If a customer does not have enough browsing data, ODP will use the best selling products from the last week instead. Use this option to remind customers of other products they have been interested in recently.
  6. Click Save.


Add to an email (Legacy)

You can add a content section based on recommendations to an email. 

  1. Go to Campaigns. 
  2. Click the name of the campaign you want to edit or create a new campaign.
  3. Select the Edit icon of the email touchpoint you want to personalize.
  4. Drag a Dynamic Grid into your email from Elements > Layouts
  5. Select the default recommendation or custom behavior from Source
  6. Click Save.