Einstein Carousels: when and how to use them to get the most out of SFCC

Beyond the PDP: the strategic positioning of carousels

According to a Baymard Institute benchmark, guiding users by showing both alternative and complementary products is crucial for conversion within a Product Page. Yet it has emerged that 58% of e-commerce sites fail to leverage this opportunity, mixing these logics or using only one of them.

Implementing a carousel with clearly distinct objectives within the Product Page is therefore the starting point, but the most serious mistake is limiting Einstein’s Artificial Intelligence exclusively to this page.

The real leap in quality, in fact, occurs when the carousel goes beyond the boundaries of the PDP to become a fluid and strategic element throughout the entire funnel.

1. 1. Homepage
   A place that can evolve from an exploratory showcase to a conversion shortcut.
   For a new visitor the Homepage acts as a discovery catalog, but for a returning user it must transform into a fast-track to checkout.

• Recommended Strategy: Replace generic modules with an Einstein carousel based on the Recently Viewed logic, welcoming the user from their very first landing with the products that have already captured their attention.
• UX Impact: Instead of forcing the user to navigate again through menus and filters, the algorithm re-engages the purchase intent from the previous session, reducing friction and speeding up navigation.

2. Cart Page: the kingdom of upselling and AOV
The cart is the stage at which purchase intent reaches its peak. Yet, according to a Baymard Institute benchmark, as many as 52% of e-commerce sites offer entirely irrelevant cross-selling suggestions at this step, or suggestions based on overly generic logic.
At this stage, Einstein’s objective must not be the discovery of new products, but a dynamic, low-friction push to increase the Average Order Value (AOV).

• Recommended Strategy: Implement a Recommender Strategy based on logics such as Bought Together or Complementary. The focus should be on suggesting related accessories or lower-value products relative to the cart contents, helping the user easily reach the free shipping threshold.
• CRO Impact: Suggesting the right item at the moment the user is about to complete payment is the most powerful upselling lever.

3. “Zero Results” and 404 Pages: turning abandonment into opportunity
An internal search that yields no results or landing on a 404 page represent the main obstacles to conversion. In these critical moments, the risk of site abandonment is at its highest.

• The Strategy: Avoid “blank pages” and implement a predictive carousel that suggests Best Sellers or Recently Viewed products.
• UX Impact: Transform a moment of strong frustration and high exit propensity into a new entry point for navigation. According to Salesforce’s global Shopper-First Retailing report, visitors who interact with a personalized experience (such as a carousel) are 4.5 times more likely to add items to their cart compared to those browsing in standard mode.

The importance of context: Einstein and Customer Groups

Showing the same carousel to every type of user means ignoring the value of segmentation. In a hyper-competitive market, generic personalization is no longer sufficient: the real leap in quality on Salesforce occurs by integrating Einstein’s intelligence with Customer Groups.

By combining these two tools, it is possible to serve dynamic recommendations not only based on browsing behavior, but also based on customer value and price sensitivity.
The goal must be to go beyond a single Recommender Strategy per page that is the same for everyone, and instead create dynamic slots that adapt to the user’s segment.

• Example for “VIP / High Spenders” users: For loyal customers with a high average order value, the lever is not price, but exclusivity. In this case, it is ideal to show carousels driven by the New Arrivals logic or high-margin premium items, intercepting their desire for novelty.
• Example for “Price Sensitive” or Guest users: For those who browse primarily in the outlet section or land on the site for the first time, the barrier to purchase is higher. Here the winning strategy is to show Best Sellers on sale or products with the most competitive price in the category, increasing the likelihood of immediate conversion.

Measurement and Testing: making Einstein a Data-Driven choice

Many brands make the mistake of evaluating Einstein’s effectiveness based solely on the native dashboard in Business Manager. While useful for an overall view, this tool is not sufficient to conduct a real optimization strategy.

Validating hypotheses: the A/B Testing approach

Since every e-commerce responds to different industries, user bases, and purchasing habits, there is no universal formula, and the only way to evaluate the best option is through constant testing.

In the Salesforce Commerce Cloud environment, the most effective way to do this is to use Content Slots, configuring two or more different Einstein strategies within the same slot. This makes it possible to split traffic and measure in real time which logic generates the greatest impact on KPIs.

Choosing strategies: some immediately testable examples

• PDP Test (Related Products vs Complete the Look): on the product page, compare a similar products logic (Related) with a cross-selling one (Complete the Look). This allows you to understand whether the user at that moment is looking for an alternative to what they are viewing, or whether they are ready to add combined products or accessories to complete the purchase.
• Cart Test (Accessories vs Lower Price): within the Cart Page, test a strategy focused on the “Accessories” category against one that dynamically shows products with a price lower than the current cart value. The latter is a very powerful lever for pushing the user toward the free shipping threshold with minimal decision-making friction.

Which KPIs to monitor at the end of an A/B Test?

To evaluate the real impact of a test on Einstein and determine the winning variant, it is essential to cross-reference engagement data with business metrics.
The essential KPIs to analyze are:

• Click-Through Rate (CTR) on carousel products: a higher CTR means that the positioning is correct and that the chosen logic is effectively capturing the user’s attention.
• Add to Cart (ATC) Rate: confirms whether the product suggested by the algorithm has not only piqued the user’s interest, but has also generated a real purchase intent.
• Conversion Rate (CR): the final conversion rate of the session demonstrates whether the interaction with that specific Einstein strategy actually facilitated the purchase.
• Average Order Value (AOV): essential, especially for cross-selling tests in the cart. The winning strategy must demonstrate its ability to dynamically increase the average order value.
• Total Revenue: this is the definitive metric that measures the overall impact on revenue generated by the individual variant, justifying the investment in the platform and the optimization of the A/B test.

Beyond SFCC data: custom tracking and ad hoc dashboards

Relying solely on Salesforce Commerce Cloud’s native dashboards provides only a partial picture of performance.
To concretely evaluate the effectiveness of carousels, it is optimal to integrate platform data with custom tracking on GA4 and dedicated dashboards.
A setup of this kind allows you to go beyond the single-click metric, measuring actual views and cross-referencing interactions with the user’s overall behavior on the site. This provides a complete and necessary view for truly optimizing the strategy.

These are just some of the best practices for optimizing Einstein carousels on Salesforce Commerce Cloud, but every e-commerce has specific needs. Thanks to our experience and expertise in UX and CRO, we can help you truly maximize conversions and AOV, transforming Einstein’s Artificial Intelligence into a real business accelerator through targeted analysis.