How to choose the right algorithm
Select the best-fit logic for each page or experience
Overview
Liva 7’s personalization engine lets you choose from a range of recommendation algorithms - each optimized for different goals, shopper behaviors, and page types. This guide helps you understand what each algorithm does and when to use it.
Step 1: Go to the Targeting rules tab
From the Liva 7 dashboard, open the Targeting rules menu item and under the Engine status tab, you’ll see:
- A toggle to turn the engine On/Off
- A dropdown labeled Algorithm - this controls the global recommendation logic
This algorithm applies storewide unless overridden by a Page Strategy or Widget-level setting.
Step 2: Explore the algorithm options
Here’s what each algorithm does and when to use it:
• Smart Mix (Recommended)
A dynamic blend of behavioral, similarity, and popularity signals. Ideal for most page types and general personalization.
• Customers Also Bought
Recommends items frequently purchased with the current product. Best for product pages and cart upsells.
• Similar Products
Shows visually or functionally similar items. Great for product pages and out-of-stock alternatives.
• Based on Browsing History
Personalizes based on what the shopper has viewed recently. Works well on the homepage or collection pages.
• Based on Purchase History
Tailors recommendations to past orders. Effective for returning customers and loyalty campaigns.
• Trending Now
Highlights popular products across your store. Use this on the homepage or collection pages to drive urgency.
• New Arrivals
Surfaces recently added products. Perfect for showcasing fresh inventory or seasonal launches.
Smart Mix adapts to different shopper behaviors and page types - it’s a great default if you’re unsure where to start.
Step 3: Match algorithm to page context
Use Smart Mix as your default if you want a balanced, adaptive approach. For more targeted use cases, align algorithms with specific page types:
• Product pages - Similar Products or Customers Also Bought
• Cart pages - Customers Also Bought or Smart Mix
• Homepage - Smart Mix, Trending Now, or Based on Browsing History
• Collection pages - Trending Now or Based on Purchase History
• Search results - Similar Products or Content-Based (if available)
You can override the global algorithm per page using Page Strategies or per widget using Widget Settings.
Step 4: Save your selection
Once you’ve chosen the algorithm that fits your goals, click Save Strategy. The engine will begin using your selected logic in real time. You can return to this setting anytime to test new approaches or respond to seasonal changes.
Making it work for you
Choosing the right algorithm is one of the most important steps in delivering relevant, high-performing recommendations. Start with Smart Mix for broad coverage, then experiment with more focused strategies as you refine your personalization setup.
