PureClarity automatically tracks all recommender interactions and attributes revenue to successful recommendations, providing clear ROI measurement for your personalization efforts.
Automated Recommender Campaigns
Recommended starting point for most stores, automated recommenders use machine learning to deliver the most relevant recommendations for each customer and context.How Automated Recommendations Work
PureClarity analyzes multiple data points to select optimal recommendations: Customer behavior factors:- Current browsing: Products being viewed and categories explored
- Basket contents: Items added but not yet purchased
- Purchase history: Previous orders and preferences
- Session patterns: Time spent, pages visited, interaction depth
- Page type: Homepage, product page, cart, category page
- Product relationships: Cross-sells, upsells, alternatives
- Inventory levels: Availability and stock status
- Seasonal trends: Current demand patterns and popularity
Page-Specific Recommendation Types
Homepage recommendations:- “Recommended based on your last visit”: Personalized for returning customers
- “Trending Products”: Popular items for new visitors
- “New Arrivals”: Fresh inventory for engaged customers
- “Best Sellers”: High-converting products for broad appeal
- “Frequently bought together”: Cross-sell opportunities
- “Customers also bought”: Alternative and complementary products
- “Similar products”: Alternatives within the same category
- “Complete the look”: Style and accessory recommendations
- “Add to your order”: Last-minute upsells
- “Others also bought”: Based on cart contents
- “Recommended for you”: Personalized additional items
Automated Recommender Settings
Minimum items: Set the minimum number of products required for display Maximum items: Control the maximum products shown per recommendation Fallback behavior: What happens when insufficient products are availableStart with 3-6 products per recommender for optimal balance between choice and decision fatigue. Adjust based on your specific product catalog and customer behavior.
Custom Recommender Campaigns
Custom recommenders provide editorial control over product selection while maintaining the dynamic nature of personalized recommendations.If no products match your custom recommender criteria for a specific customer, the campaign will not display. Always include fallback options or broader criteria.
Custom Recommender Sources
Manual Recommenders- Specific products: Hand-selected items for precise control
- Product categories: Target specific product types or collections
- Brand focus: Highlight particular brands or vendors
- Use cases: Clearance promotions, “complete the look” styling, editorial features
- Top performers: Products with highest sales volume
- Trending items: Products with increasing popularity
- Category leaders: Best sellers within specific categories
- Consideration: May overlap with automated recommendations
- AI curation: Personalized selection based on customer data
- Behavioral matching: Products aligned with customer interests
- Dynamic updating: Changes based on recent activity
- Limitation: May not have recommendations for all customers
- Browsing history: Products customer has viewed recently
- Re-engagement: Encourage return to considered products
- Session continuity: Maintain shopping journey context
- Availability: Depends on customer having browsing history
- Rising popularity: Products with highest sales velocity increase
- Market dynamics: Items gaining traction with customers
- Fresh discovery: Showcase emerging popular products
- Algorithmic selection: Based on sales rank changes
Product Filtering Options
Refine recommendations with advanced filtering criteria: Price Range Filters- Minimum price: Exclude products below specified amount
- Maximum price: Limit recommendations to affordable range
- Use cases: Budget-conscious segments, luxury targeting
- Specific categories: Limit to particular product types
- Multi-category: Include multiple related categories
- Exclusions: Remove categories not relevant to campaign
- Brand inclusion: Focus on specific brands or vendors
- Brand exclusion: Remove particular brands from recommendations
- Partnership focus: Highlight partner or featured brands
- Product tags: Use platform-specific product tagging
- Custom attributes: Filter by custom product properties
- Seasonal tags: Include/exclude seasonal or promotional items
- Platform dependency: Implementation varies by e-commerce system
Contact your customer success manager for guidance on implementing tag-based filtering specific to your e-commerce platform.
Custom Recommender Settings
Campaign title: Customize the display title for the recommendation section Minimum/Maximum items: Control quantity displayed Fallback strategy: Define behavior when criteria yield insufficient productsPerformance Optimization Strategies
Automated vs. Custom Decision Framework
Choose Automated when:- Starting with personalization
- Seeking maximum conversion optimization
- Limited time for content curation
- Wanting set-and-forget functionality
- Promoting specific products or brands
- Running targeted promotional campaigns
- Need editorial control over recommendations
- Testing specific product combinations
A/B Testing Recommendations
Test different approaches:- Automated vs. Custom: Compare performance in similar zones
- Product quantities: Test different min/max settings
- Filter combinations: Experiment with different criteria
- Title variations: Test different recommendation headers
Common Optimization Tactics
Underperforming recommenders:- Broaden criteria: Reduce restrictive filters
- Adjust quantities: Test different product counts
- Review placement: Consider zone location and timing
- Audience refinement: Narrow or broaden target segments
- Scale successful patterns: Apply winning formulas to other zones
- Refine further: Test minor adjustments for additional gains
- Expand audience: Gradually broaden successful targeting
- Document learnings: Capture insights for future campaigns
Attribution and Revenue Tracking
Automatic Click Tracking
PureClarity automatically tracks:- Product clicks: When customers interact with recommended products
- Purchase attribution: Sales resulting from recommendation clicks
- Revenue totals: Dollar value attributed to specific recommendations
- Conversion rates: Success rates for different recommendation types
Click Total Analytics
Zone-level reporting: Revenue attribution by website location Campaign-level insights: Performance of specific recommendation campaigns Product-level data: Which products perform best in recommendations ROI analytics: Detailed return on investment for recommendation effortsAccess comprehensive recommendation performance data through the Analytics section and specialized Recommender ROI reports.
Attribution Best Practices
Revenue tracking accuracy:- Consistent attribution windows: Define clear timeframes for conversion credit
- Multi-touch attribution: Understand customer journey complexity
- Baseline comparison: Measure lift versus non-personalized experiences
- Regular review cycles: Weekly/monthly performance assessment
- Trend analysis: Identify patterns and seasonal variations
- Segment analysis: Compare performance across customer groups