NOTE: Product Recommendations is a paid add-on to Zaius. If you do not have access to Product Recommendations in your account, please reach out to your Customer Success Manager. 


Personalized product recommendations are a powerful tool for driving sales, increasing conversion rate, and increasing order value. Savvy brands use dynamic product content to personalize the offers that are delivered to each customer. 

In Zaius, you can dynamically change entire sections of content within emails to make campaigns more relevant and appealing to customers. These product recommendations can be included in newsletters, abandon cart/browse campaigns, lifecycle nudge campaigns, etc. Zaius provides four default options for product recommendations: 

  • Zaius Recommendations: Automated recommendations based on each user's behavior compared to other users with similar behaviors (see more below)
  • Best Selling Products: Products that are purchased most 
  • Most Viewed: Products that are viewed most 
  • Revenue Generators: Products that produce the most revenue

About Zaius Recommendations

Zaius Recommendations employs machine learning to analyze the purchase history of each customer to make smart, data-driven predictions about what they may want to buy in the future. Zaius detects purchasing patterns in your ecommerce data and uses them to automatically predict your customers’ buying behavior, so you can target the right people with the right products. Recommendations are determined based on each user's behavior compared to other users with similar behaviors. The machine learning algorithm runs nightly using event data from the previous 180 days. 

Ideal Campaigns 

When populating product recommendations in ZED, Zaius utilizes the behavior of each of your customers, compared to that of other customers, to determine dynamic, personalized recommendations. Given the importance of behavioral data for Zaius Recommendations, this feature is ideal for:

  • Campaigns targeting repeat purchasers, as it is more likely that we have a long history
  • Newsletters, as most customers will have a longer history
  • Welcome campaigns, as we may have enough history to generate recommendations if we have a history of anonymous data before signup/identification

Zaius Recommendations are not ideal for:

  • Winback campaigns, as you won't have any information on customers if they haven't been seen in 180 days (NOTE: Winback campaigns that run after 3 months of inactivity could still be effective)
  • Clients with small catalogues, as customers may get the same recommendations
  • Clients with seasonal releases, as customers may get recommendations that are no longer applicable

Zaius Recommendations will be effective for some customers for specific campaigns. However, if you have a solid understanding of your customers, you may be able to better target your customers without the use of machine learning.

Default Recommendations 

If a particular customer does not yet have enough behavioral data, Zaius will use default recommendations instead. Those default recommendations are defined by the best selling products from the last week.

If you see recommendations repeated for several of your customers, you are likely seeing default recommendations. But don't fret! As Zaius collects additional behavioral data about your customers, the recommendations will become further tailored for your audience.

Creating a Dynamic Grid with Recommendations 

To create a Dynamic Grid based on Recommendations, you must first create the Dynamic Grid. Once the Dynamic Grid has been added, click on the Details tab of the right panel.

  1. Select Products from the Feed dropdown.
  2. Select any of the four Recommendations options from the Source dropdown.
  3. Select the Timeframe. 

Next, pull Elements from the right pane into your Dynamic Grid and populate the necessary fields using Liquid. For example, typical setup would contain:

  • Image element with image URL: {{product.image_url}}
  • Text element containing product name: {{}}

Note that whichever Elements you include in the first cell will be automatically repeated for other cells. 

Finally, preview the email to make sure it is working correctly. 

Did this answer your question?