Analytics and AI based Retention in e-commerce
Download as PPTX, PDF3 likes1,064 views
The document discusses the strategies for optimizing customer retention in e-commerce through analytics and AI-based solutions. It emphasizes the importance of user segmentation, engagement experiences, and data-driven decision-making to enhance customer experiences and reduce churn. Additionally, it highlights the role of predictive modeling and content recommendation engines in improving marketing effectiveness and user retention rates.
Analytics and AI based Retention in e-commerce
- 1. Analytics and AI based Retention in e-commerce Jay Magdani
- 2. We help Optimize Customer Experiences at Scale Customer Lifecycle Management and Engagement Platform Cloud Based Digital Marketer AARRR 10B+ App and Web Analytics Engagement 26M/min CleverTap?
- 3. Analyst Retention: How e-commerce plays the game ATTRACT CONVERTRETAIN MarketerGrowth Product
- 4. Day in a life of an e-commerce business Strategic Landscape Retention: What the industry thinks App performance? How do I improve revenue? How do I get stickiness? Who is my User? How do I reduce Churn? What should I build next? Tactical Landscape Who should I reach out to? Why are my users exhibiting said behavior? Where should I target my Users? What should be my message? When is the best time to reach my users
- 5. Martech Landscape Day in a life of a marketer Analytics Landscape Make smarter decisions Retention: What e-commerce demands from technology How does my App perform? How do I meet my KPIs? Who? Why? When? Where?What? Descriptive Analytics Prescriptive Analytics Predictive Analytics Diagnostic Analytics C L E V E R T A P
- 6. Retention: What e-commerce demands from technology SEGMENTATION ANALYTICS ENGAGEMENT EXPERIENCES Behavioral | Interest | Demographic | Geographic | RFM | Intent Funnels | Cohorts | Flows Push | Email | WhatsApp Experiments | Triggers
- 7. But before that… Collect data Profile • Information about the user • Geo, Demo, type, devices Event • Actions performed by user • App Launch, Charged Who I am? End User Profile Event Engage Analytics What I do? Brand
- 8. AUTOMATED USER SEGMENTATION CHAMPION USER OPTIMIZATION • Analyze what the champion users do • Track and compare behavior with other users • Replicate the champion user behavior 1 2 How AI plays a role… SEGMENTATION ANALYTICS ENGAGEMENT EXPERIENCES CONTENT RECOMMENDATION • What message should I send? • Performance of campaigns based on message • Track, measure and learn from ROI 3 OPTIMIZE NATIVE UI EXPERIENCES • Blue button vs Red button • Run experiments • Easily publish multiple app versions 4
- 9. History with Segmentation Manual Segmentation Tool Automated Segmentation AUTOMATED USER SEGMENTATION 1 Assisted Segmentation Via RFM
- 10. Manual Segmentation Problems AUTOMATED USER SEGMENTATION 1 Intuition Based Best users? Users who did App Launch and then… Searched and then… Added to Cart and then… Charged Within 1 day
- 11. RFM Problems AUTOMATED USER SEGMENTATION 1 Single Event Analysis Ignores Time Aspect Low Predictive Power Reactive Split all users who made a Purchase into good and bad segments
- 12. Why Intent Based Segmentation AUTOMATED USER SEGMENTATION Proactive 1 Users who will uninstall in the next 30 days Users Most Likely Moderately Likely Least Likely
- 13. Productizing Predictive Segments AUTOMATED USER SEGMENTATION 1
- 14. Largest ticket booking app in LATAM ● Before ○ High uninstall rates ○ Reactive approach ○ Low conversions Large food app in Asia ● Before ○ High marketing spend Impact on Customers AUTOMATED USER SEGMENTATION 1 IBS Impact Conversions 25% Uninstalls 18% Spend 21% ● After ● After ○ Optimized budget by selective, smart and intent based engagement strategy
- 15. AUTOMATED USER SEGMENTATION • Real-time User Segmentation • Customer Insights powered by ML • Predictive modeling for better ROI CHAMPION USER OPTIMIZATION 1 2 SEGMENTATION ANALYTICS ENGAGEMENT EXPERIENCES CONTENT RECOMMENDATION • What message should I send? • Performance of campaigns based on message • Track, measure and learn from ROI 3 OPTIMIZE NATIVE UI EXPERIENCES • Blue button vs Red button • Run experiments • Easily publish multiple app versions 4
- 16. Compare Conversion Funnels CHAMPION USER OPTIMIZATION 2 App Launched Product Viewed Added to Cart Charged App Launched Product Viewed Added to Cart Charged 1M 100K 500K 100K 25K 25K 10K 1K Champion Users New Users
- 17. CLTV CHAMPION USER OPTIMIZATION 2 Customer Acquisition CostTotal Revenue generated Duration before churn E-commerce Golden Ratio -> 3:1
- 18. AUTOMATED USER SEGMENTATION • Real-time User Segmentation • Customer Insights powered by ML • Predictive modeling for better ROI CHAMPION USER OPTIMIZATION 1 SEGMENTATION ANALYTICS ENGAGEMENT EXPERIENCES CONTENT RECOMMENDATION OPTIMIZE NATIVE UI EXPERIENCES • Blue button vs Red button • Run experiments • Easily publish multiple app versions 4 3 2 • Analyze what the champion users do • Track and compare behavior with other users • Replicate the champion user behavior
- 19. ● Poor user experience ● High churn rates ● Lost business opportunity ● Low LTV/CLV CONTENT RECOMMENDATION 3 ● Category Based ● Same recommendations provided multiple times. REPEAT RECOMMENDATIONS ● Using basic analytics like best items, etc. ● Best guess intuition MANUAL DISCOVERY Issues with Manual Content Creation
- 20. CONTENT RECOMMENDATION 3 Why Content Recommendation Engine Right Content Vestibulum congue Vestibulum congue Vestibulum congue Vestibulum congue Relevance SerendipityDiversity Right User
- 21. CONTENT RECOMMENDATION 3 How recommendation engine works Upload Catalog Run Recommendation Model Engage Luke: Views electronics goods John: Searches sports goods Product Name Category Price iPhone 11 Electronics $ 700 Tennis shoes Sports $ 400
- 22. ● Using Recommendations in Campaigns and Journeys ● Comparing performance of recommendations w.r.t Control Group ● CTR and Conversion performance Boost generated CONTENT RECOMMENDATION 3 Impact on Customers 40% ARPU
- 23. AUTOMATED USER SEGMENTATION • Real-time User Segmentation • Customer Insights powered by ML • Predictive modeling for better ROI CHAMPION USER OPTIMIZATION 1 SEGMENTATION ANALYTICS ENGAGEMENT EXPERIENCES CONTENT RECOMMENDATION OPTIMIZE NATIVE UI EXPERIENCES 2 • Analyze what the champion users do • Track and compare behavior with other users • Replicate the champion user behavior 4 3 • What message should I send? • Performance of campaigns based on message • Track, measure and learn from ROI
- 24. OPTIMIZE NATIVE UI EXPERIENCES 4 What is it?
- 25. OPTIMIZE NATIVE UI EXPERIENCES 4 What is it? Purchase Now!!! Offer available 30% off Last day today Buy 30% discount
- 26. OPTIMIZE NATIVE UI EXPERIENCES 4 Why is it important? money time experience optimization
- 27. Prescriptive Recommenders Predictive IBS Diagnostic Segment Comparison, Anomaly Detection and Reasons, etc. Descriptive Events, Trends, RFM, Flows, etc. Data Science Pillars Summary: Driving Retention backed by Analytics and DS