Customer Intelligence & Analytics - Part IV: An Evolutionary Process: Moving Toward Analytically Driven Marketing
2 likes665 views
The document discusses different approaches to marketing optimization including campaign prioritization, customer rules, and full optimization. It shows that full optimization, which considers all business objectives, constraints, and allocation options simultaneously, typically yields the highest expected return compared to other approaches that make incremental decisions or do not consider all factors. Real-world examples are provided to illustrate the benefits of optimization over simpler approaches.
Customer Intelligence & Analytics - Part IV: An Evolutionary Process: Moving Toward Analytically Driven Marketing
- 2. Module 4: An Evolutionary Process - Moving Toward Analytically Driven Marketing 3.1 Introduction 3.2 Marketing optimization 3.3 The art and science of the marketing mix 3.4 Real-world, success case studies 3.5 Questions
- 3. • Debbie Mayville – Sr. Solutions Architect, Communications & Marketing Analytics, SAS • David Kelley – Sr. Solutions Architect, Customer Intelligence, SAS • Suneel Grover – Solutions Architect, Integrated Marketing Analytics, SAS – Adjunct Professor, Integrated Marketing Analytics, New York University (NYU)
- 4. Module 4: An Evolutionary Process - Moving Toward Analytically Driven Marketing 3.1 Introduction 3.2 Marketing optimization 3.3 The art and science of the marketing mix 3.4 Real-world, success case studies 3.5 Questions
- 5. The Marketing Process Mobile Online Finance Risk Call Customer Center Service In Person Merchandising Social Corporate Affairs Direct Mail Marketing Operations Optimization Marketing Marketing Marketing Strategy Processes Campaigns Analytics Data Integration ERP CRM EDW Online Social Campaign
- 6. The Marketing Process Mobile Online Finance Risk Call Customer Center Service In Person Merchandising Social Corporate Affairs Direct Mail Marketing Operations Marketing Mix Real-Time Campaign Optimization Management Analysis Decisioning Marketing Marketing Performance Operations Online Customer Social Media Management Management Behaviour Analytics Data Integration ERP CRM EDW Online Social Campaign
- 7. Optimization Defined Optimization A computational problem in which the objective is to find the best of all feasible solutions
- 8. The Relationship Marketing Context • Many customers, offers, channels • Managing the contact strategy • Looking ahead and behind • How do you allocate offers effectively to maximize return? • Many constraints impact decisions Budgets, resources, policies • How to respect constraints? • How to reconcile competing goals? • How to plan effectively for change?
- 9. Marketing Optimization Marketing Optimization “What should I do to achieve the best results?“ Marketing Simulation “What would happen?" Business Value Predictive Modeling “How likely are my customers to respond to an offer?” Marketing Dashboard “How many new customers did we get last Data Quality, month? How much customer attrition?" Integration “How can we trust analysis if we don’t trust the data?” Data Access “What measures are available to better understand our business?” Reactive Proactive Predictive Strategic Intelligence
- 10. Massive Problem - Potential Choices Product A Product B Product C
- 11. Marketing Optimization Applications • Financial Services – Insurance policy offers – Credit line increase/decrease – APR to offer on balance transfer offers • Telecom – Complex cell phone plan offers – Bundled services – Cross channel offers with different execution costs • Hospitality (Hotels, Casinos) • Loyalty offers • Retail • Personalized coupons (POS) • Offer prioritization and collisions • Contact stream optimization
- 12. Do All Marketing Approaches Yield The Same Results? 10–100+ % Optimization - Solves by holistic 5-10 % approach - Factors all constraints Customer Rules - Determines the best result - First In, First Out ? - Prioritized by Customer/Campaign Prioritization - Fails in the face of constraints - First In, First Out - Prioritized by Campaign - Does not provide best combination
- 13. Optimization Techniques Example • Lines of business = 3 • Return = expected value (probability*expected revenue) • Business objective = maximise value • Constraints: Each customer is assigned to at most 1 campaign Each campaign can have at most 3 customers Client Camp’ A Camp’ B Camp’ C 1 100 120 90 2 50 70 75 Campaign C 3 60 75 65 4 55 80 75 5 75 60 50 6 75 65 60 7 80 70 75 Campaign B Campaign A 8 65 60 60 9 80 110 75
- 14. Optimization Techniques – Campaign Prioritization • Campaigns assigned a priority • Customers allocated to campaigns by expected customer value Client Camp’ A Camp’ B Camp’ C 1 100 120 90 2 50 70 75 3 60 75 65 Campaign C 4 55 80 75 5 75 60 50 6 75 65 60 7 80 70 75 Campaign B Campaign A 8 65 60 60 9 80 110 75
- 15. Cross-channel Optimisation Campaign Prioritization Constraints: Expected Return: 260 ??? 1 customer - 1 campaign 1 campaign - 3 customers Client Camp’ A Camp’ B Camp’ C 1 100 120 90 2 50 70 75 Campaign C 3 60 75 65 4 55 80 75 5 75 60 50 6 75 65 60 Campaign B Campaign A 7 80 70 75 8 65 60 60 9 80 110 75
- 16. Campaign Prioritization Constraints: Expected 485 260 1 customer - 1 campaign Return: 1 campaign - 3 customers Client Camp’ A Camp’ B Camp’ C 1 100 120 90 2 50 70 75 Campaign C 3 60 75 65 4 55 80 75 5 75 60 50 6 75 65 60 Campaign B Campaign A 7 80 70 75 8 65 60 60 9 80 110 75
- 17. Campaign Prioritization Constraints: Expected Return: 655 485 1 customer - 1 campaign 1 campaign - 3 customers Client Camp’ A Camp’ B Camp’ C 1 100 120 90 2 50 70 75 Campaign C 3 60 75 65 4 55 80 75 5 75 60 50 6 75 65 60 Campaign B Campaign A 7 80 70 75 8 65 60 60 9 80 110 75
- 18. Optimization Techniques - Customer Rules • Customers assigned a priority • Campaigns allocated to customers by expected customer value Client Camp’ A Camp’ B Camp’ C 1 100 120 90 2 50 70 75 Campaign C 3 60 75 65 4 55 80 75 5 75 60 50 6 75 65 60 Campaign B Campaign A 7 80 70 75 8 65 60 60 9 80 110 75
- 19. Customer Rules Constraints: Expected Return: 120 ??? 1 customer - 1 campaign 1 campaign - 3 customers Client Camp’ A Camp’ B Camp’ C 1 100 120 90 2 50 70 75 Campaign C 3 60 75 65 4 55 80 75 5 75 60 50 6 75 65 60 Campaign B Campaign A 7 80 70 75 8 65 60 60 9 80 110 75
- 20. Customer Rules Constraints: Expected Return: 195 1 customer - 1 campaign 1 campaign - 3 customers Client Camp’ A Camp’ B Camp’ C 1 100 120 90 2 50 70 75 Campaign C 3 60 75 65 4 55 80 75 5 75 60 50 6 75 65 60 7 80 70 75 Campaign B Campaign A 8 65 60 60 9 80 110 75
- 21. Customer Rules Constraints: Expected Return: 195 270 1 customer - 1 campaign 1 campaign - 3 customers Client Camp’ A Camp’ B Camp’ C 1 100 120 90 2 50 70 75 Campaign C 3 60 75 65 4 55 80 75 5 75 60 50 6 75 65 60 Campaign B Campaign A 7 80 70 75 8 65 60 60 9 80 110 75
- 22. Customer Rules Constraints: Expected Return: 270 350 1 customer - 1 campaign 1 campaign - 3 customers Client Camp’ A Camp’ B Camp’ C 1 100 120 90 2 50 70 75 Campaign C 3 60 75 65 4 55 80 75 5 75 60 50 6 75 65 60 7 80 70 75 Campaign B Campaign A 8 65 60 60 9 80 110 75
- 23. Customer Rules Constraints: Expected Return: 425 1 customer - 1 campaign 1 campaign - 3 customers Client Camp’ A Camp’ B Camp’ C 1 100 120 90 2 50 70 75 Campaign C 3 60 75 65 4 55 80 75 5 75 60 50 6 75 65 60 7 80 70 75 Campaign B Campaign A 8 65 60 60 9 80 110 75
- 24. Customer Rules Constraints: Expected Return: 500 425 1 customer - 1 campaign 1 campaign - 3 customers Client Camp’ A Camp’ B Camp’ C 1 100 120 90 2 50 70 75 Campaign C 3 60 75 65 4 55 80 75 5 75 60 50 6 75 65 60 7 80 70 75 Campaign B Campaign A 8 65 60 60 9 80 110 75
- 25. Customer Rules Constraints: Expected Return: 500 580 1 customer - 1 campaign 1 campaign - 3 customers Client Camp’ A Camp’ B Camp’ C 1 100 120 90 2 50 70 75 Campaign C 3 60 75 65 4 55 80 75 5 75 60 50 6 75 65 60 7 80 70 75 Campaign B Campaign A 8 65 60 60 9 80 110 75
- 26. Customer Rules Constraints: Expected Return: 640 1 customer - 1 campaign 1 campaign - 3 customers Client Camp’ A Camp’ B Camp’ C 1 100 120 90 2 50 70 75 Campaign C 3 60 75 65 4 55 80 75 5 75 60 50 6 75 65 60 7 80 70 75 Campaign B Campaign A 8 65 60 60 9 80 110 75
- 27. Customer Rules Constraints: Expected Return: 715 +60 1 customer - 1 campaign 1 campaign - 3 customers Client Camp’ A Camp’ B Camp’ C 1 100 120 90 2 50 70 75 Campaign C 3 60 75 65 4 55 80 75 5 75 60 50 6 75 65 60 7 80 70 75 Campaign B Campaign A 8 65 60 60 9 80 110 75
- 28. Optimization Techniques - Optimization Business objectives, constraints, contact policies define ‘priority’ Optimization decides allocation Client Camp’ A Camp’ B Camp’ C 1 100 120 90 2 50 70 75 3 60 75 65 Campaign C 4 55 80 75 5 75 60 50 6 75 65 60 7 80 70 75 Campaign B Campaign A 8 65 60 60 9 80 110 75
- 29. Optimization Constraints: Expected Return: 745 +30 1 customer - 1 campaign 1 campaign - 3 customers Client Camp’ A Camp’ B Camp’ C 1 100 120 90 2 50 70 75 Campaign C 3 60 75 65 4 55 80 75 5 75 60 50 6 75 65 60 7 80 70 75 Campaign B Campaign A 8 65 60 60 9 80 110 75
- 30. Marketing Optimization: Process Flow Planned Campaigns Eligible Customers Marketing Optimization Identify & Define Review Model Execute Optimization Optimization Scores Optimal Scenarios Results Outcome Contact “What-If Analysis” Policy Optimization Parameters: • Objective • Suppression Rules • Constraints: • Budget • Capacity • Contact / Blocking Policies
- 31. Case Study: Commerzbank, Germany Challenges Business Impact • 4 million customers, 20 offer types • POV: Up to 80% ROI improvement • Optimize utilization of consultants • Production: 50% yield with the same budget • Optimize Yield vs. Budget • ROI increased by 407% • Optimize Marketing ROI (revenue / cost) "We have compared SAS intensively with other manufacturers offerings. The result was impressive: SAS Marketing Optimization is exactly the solution we were looking for. We are +407% setting an industry Benchmark” ROI Heiko Güthenke, Department Director Customer & Business Analysis
- 32. More Case Studies… Client Name Benefits Vodafone (Australia) • 3-10x Response Rate increase • Improve campaign ROI by 4x • 30% reduction in campaign costs Scotiabank • 50% Campaign ROI improvement Major Insurer • 12% increase in revenue; 52% in earnings • Savings of >$4 million per year U.S. Regional Telco • $6 million incremental LTV in the 1st month Global Telco • Reduced call center contacts by 25% without decreasing effectiveness #1 Market Share European • Individualized targeting of monthly coupon Retailer mailers • Increased offer response rates • Decrease mailing costs
- 33. Module 4: An Evolutionary Process - Moving Toward Analytically Driven Marketing 3.1 Introduction 3.2 Marketing optimization 3.3 The art and science of the marketing mix 3.4 Real-world, success case studies 3.5 Questions
- 34. The Marketing Process Mobile Online Finance Risk Call Customer Center Service In Person Merchandising Social Corporate Affairs Direct Mail Marketing Operations Marketing Mix Real-Time Campaign Optimization Management Analysis Decisioning Marketing Marketing Performance Operations Online Customer Social Media Management Management Behaviour Analytics Data Integration ERP CRM EDW Online Social Campaign
- 35. Increased Complexity With Marketing How do you decide the right mix across all channels? Web Web Social Media & Direct Word of Customer Sales (Corp) (eCommerce) Media Ads Mail Mouth Service Advertising Email & Social Retail Interactive Direct 1:1 & Mobile Marketing Marketing Promotions
- 36. Above The Line…Below The Line… » Above the Line » Below the Line
- 37. Above the Line…Below the Line… Media Planner/Buyer • How did we perform across products, geographies, campaign types? Brand Manager • What marketing activities drove our new sales? • What if we move funds from traditional to online marketing? Interactive Marketing • What actions/decisions to we make for various scenarios? • MOST of my marketing data is in silos…can I leverage it for analysis? Marketing Planning » Above the Line » Below the Line IT • How can I get the right offer, to the right person via the right channel? Interactive Marketing • Can I coordinate my multi-channel campaign efforts? Director of Database Mkt • Can I be relevant with EVERY interaction, every time? Campaign Planner/Designer Marketing Operations
- 38. Marketing Challenge: Financial Pressures • Aggressive corporate goals & objectives • Increased accountability and scrutiny into marketing budgets • Reductions in budgets
- 39. Questions Marketing Mix can Address • How can I still achieve my marketing goals while facing budget cuts? • I am below target, how do I re-allocate my marketing budget to hit targets? • How do I decide where to invest my marketing budget to support a product portfolio? • How and where do I invest in social media to maximize business impacts? • Where do I increase marketing investments to achieve higher returns?
- 40. What is Marketing Mix Modeling? A data driven analytic process that quantifies the relationship between drivers/influencers of sales and the resulting sales across channels • Understand the past performance of sales & marketing activities • Analyze and assess average ROI and marginal ROI • Evaluate marketing investment among ever increasing media options • Compare and assess different future marketing spending plans
- 41. Marketing Mix Technology Analytic Dashboards
- 42. Technology Capabilities Analytic dashboards • Analytic data warehouse surfaced through interactive dashboards • All media and promotions display in one location with prebuilt reports delivering summary and Analytic Dashboards detailed results Powerful analytic tools • Understand the impact of advertising on sales and incorporate into response models Adstock Analysis • Ability to explore product interactions to understand and uncover halo and cannibalization effects across your product portfolio Halo / Cannibalization Analysis
- 43. Technology Capabilities Econometric response models • Build and test time series and causal models Elasticity reports Response Model Diagnostics • Objectively quantify the relative responsiveness of each driver of sales • Decompose sales into its various components. Diminishing returns • Capture changes in marginal ROI as spending Elasticity Reports levels increase through diminishing returns curve for each channel • Determine the threshold point beyond which Sensitivity Report marketing expenditures would not yield any additional benefits Diminishing Returns
- 44. Technology Capabilities What‐if analysis & scenario planning • Ability to simulate expenditures over different Report Dashboard media and analyze the impact on products/brands/channels/geo’s Simulate/Forecast • Compare competing spending plans to understand the differences in sales Marketing mix optimization Decomposition Reports • Optimal media expense allocation for selected p roduct, channel & geography combination over a defined period of time. • Define different sets of business constraints to Compare scenarios explore the impact on the optimal solution Marketing Mix Analytics Optimization “Leave less up to chance and make data driven, evidence-based decisions”
- 45. Case Study: Large Insurance Company Business Issue Solution Benefits Quantify effectiveness of Marketing mix analytics Even though they are all marketing mix elements allows them to share consistently outspent by • Direct-response assumptions about their competitors, became • TV marketing analysis across more competitive by • Direct marketing all types of marketing determining which media • Web marketing and channels worked the • Retail channel Data is integrated from best across products and communications multiple sources and regions. analyzed to ensure accurate short-term and long-term forecasts across marketing and operations “The technology help us develop a “strategic” tool that enables us to lower risk in decision-making as we integrate all marketing disciplines with an eye toward better forecasting, budgeting, and collaboration.” Director of Strategy
- 46. Module 4: An Evolutionary Process - Moving Toward Analytically Driven Marketing 3.1 Introduction 3.2 Marketing optimization 3.3 The art and science of the marketing mix 3.4 Real-world, success case studies 3.5 Questions