Every Smart Recommender campaign includes a Recommendation Strategy, where you decide which products customers will actually see.
The primary motivation here is to show the right product, to the right person, at the right time.
Great recommendations aren’t about showing more products — they’re about showing the most relevant ones.
The Recommendation Strategy is the bridge between your catalog and your customers. It’s where data meets creativity, allowing you to combine algorithms, filters, and business logic to deliver recommendations that are both meaningful to the shopper and beneficial to your brand.
To see what it looks like in practice, let’s look at real-world-inspired examples:
Scenario 1: Cross-selling based on product attributes
Instead of relying only on the Viewed Together algorithm, Alex, a marketing manager for a home electronics brand, added filters to recommend products with complementary attributes.
Example: When a shopper viewed a smartphone, recommendations highlighted “compatible chargers,” filtered by category and a custom attribute called compatible_device.
Why it worked
Attribute-based logic avoided random upsells. The recommendations felt precise, practical, and directly relevant, which boosted accessory attachment rates.
Scenario 2: Using Manual Merchandising for seasonal control
Tom, an ecommerce dead at a gardening store, wanted to highlight seasonal products like spring gardening tools. He used the Manual Merchandising algorithm to handpick a small product set for homepage visibility, while allowing dynamic algorithms to run elsewhere.
Why it worked
Manual Merchandising provided Tom with editorial control for seasonal storytelling, without sacrificing automation. He balanced business priorities with personalization.
Scenario 3: Personalizing by User Behavior
Nadia, a campaign manager at a travel accessories brand, used the Real-Time User Engagement algorithm to show dynamically changing recommendations based on each visitor’s browsing and click activity.
Her product feed updated instantly when a user engaged with a new product category, switching from Luggages to Travel Essentials within the same session.
Why it worked
The campaign adapted like a conversation, matching the user’s intent moment by moment. This real-time responsiveness kept users engaged longer and increased conversions.
Quick Takeaways
Start with purpose, not catalog size.
Don’t show everything. Use filters and dynamic operators to keep recommendations focused and relevant.
Leverage product attributes for more brilliant storytelling.
Attributes like “compatible with,” “same color,” or “from the same collection” transform raw data into curated experiences.
Keep it fresh and available.
Outdated or irrelevant products break trust. Utilize real-time algorithms to ensure that recommendations are always based on current browsing and stock levels.