Conquer the Interwebs: The Three Pillars of Web Analytics
Download as PPTX, PDF4 likes1,050 views
Andre Rand is a web analytics manager who oversees analytics for Interline Brands, a top 100 internet retailer. He has several certifications in analytics platforms like Adobe and Google. Rand advocates for adopting Avinash Kaushik's "three pillars" approach to web analytics: behavior analysis, outcomes analysis, and experience analysis. At Interline Brands, he employs advanced strategies like RFM segmentation, incremental revenue measurement, and predictive analytics.
Conquer the Interwebs: The Three Pillars of Web Analytics
- 2. ANDRE RAND INTRODUCTION THREE PILLARS OF WEB ANALYTICS ADVANCED STRATEGIES @ INTERLINE BRANDS
- 3. ANDRE RAND INTRODUCTION Web Analytics Manager Interline Brands (Top 100 Internet Retailer) President American Marketing Association, Jacksonville Chapter Founder Creative Genius Mentorship Program for Underrepresented Minorities
- 4. ANDRE RAND INTRODUCTION (CONTINUED) Certifications • Adobe Certified Expert: SiteCatalyst Implementation • Adobe Certified Expert: SiteCatalyst User Interface • Google Analytics: Individual Qualification (GAIQ) • Silverpop Engage: Fundamental Certification • Online Testing | Landing Page Optimization | Email Marketing (MECLABS) Recent Achievements • Silverpop "Pop Star" Award for Most Dramatic Email Marketing ROI (420% -- $2.2MM) • Silverpop Customer Advisory Council Member • Interline Brands "Marketing Rock Star" Award • American Marketing Association Membership Retention Award • American Marketing Association Membership Acquisition Award
- 5. ANDRE RAND INTRODUCTION (CONTINUED) Web Analytics Objectives • Maximize ecommerce revenue and customer acquisition, retention, and reactivation for Interline Brands, a Top 100 Internet Retailer with 15 distinct industrial wholesale brands and over $2BN in annual revenue. • Analyze mountains of qualitative data (customer surveys), quantitative data (ecommerce operations, email marketing, financial reports) and predictive data (Loyalty Builders) to develop meaningful metrics and reports. • Deliver actionable recommendations to marketing, ecommerce, and IT leadership. Build web analytics infrastructure. • Drive continuous testing, learning, and evolving across 15 e-commerce websites and numerous online marketing campaigns.
- 6. ANDRE RAND INTRODUCTION (CONTINUED) Tools of The Trade Web Analytics: Multivariate Testing: Voice of the Customer: Email Marketing: Data Warehouse: Presentation: Adobe SiteCatalyst, Google Analytics Microsoft Word, Excel, PowerPoint Adobe Target, Adobe Recommendations ForeSee Results, Net Promoter Score (NPS) Silverpop Engage, Silverpop Transact Enterprise Resource Planning Systems (ERP) > Microsoft SQL Server > Access 2010
- 9. HOW TO LEARN WEB ANALYTICS? A) Read Avinash Kaushik’s Blog…Every Day • Digital Marketing Evangelist for Google and the Chief Education Officer at Market Motive Inc • Author of the leading web analytics and research blog Occam's Razor • Author of the best selling books Web Analytics: An Hour A Day and Web Analytics 2.0 • Frequent speaker at key industry conferences around the globe and at leading American universities • Developer of the “Trinity Model” aka “The Three Pillars of Web Analytics” www.kaushik.net/avinash
- 10. HOW TO LEARN WEB ANALYTICS (CONTINUED)? B) Read These Platform-Specific Books The Adobe SiteCatalyst Handbook: An Insider’s Guide Advanced Web Metrics with Google Analytics (Third Edition) C) Play in the Real World! • Use your personal website as a “sandbox” for Google Analytics, AdWords, multivariate testing, email campaign tracking, social media measurement • Adopt a “data-driven marketing” philosophy whether or not web analytics is your full-time job
- 11. HOW TO LEARN WEB ANALYTICS (CONTINUED)? D) Join the Analysis Exchange (If You’re Ready) www.webanalyticsdemystified.com/ae The Analysis Exchange is an effort designed to provide hands-on training opportunities for aspiring web analytics professionals while providing free web data analysis to the entire nonprofit community. • Organizations include nonprofits, and non-governmental organizations who rarely benefit from web analytics but have a tremendous opportunity • Students are motivated and intelligent individuals who are looking for hands- on experience with web analytics systems and real business questions • Mentors are web analytics experts willing to pledge their time to help nonprofits and students around the world
- 12. THREE PILLARS OF DATA-DRIVEN MARKETING 1. BEHAVIOR ANALYSIS 2. OUTCOMES ANALYSIS 3. EXPERIENCE ANALYSIS Source: The “Trinity” model developed by Avinash Kaushik, Digital Marketing Evangelist, Google
- 13. BEHAVIOR ANALYSIS1 The goal of behavior analysis is to infer the intent of your online customers as accurately as possible using clickstream data. “ “
- 14. BEHAVIOR ANALYSIS (CONTINUED)1 Clickstream data is the series of mouse clicks users made before, during, and after they visit our website. Traditional metrics include unique visitors, visits, page views, bounce rate, time on site, path views, and geo-location. Kick it Up a Notch To drive significant value through web analytics, you must move beyond traditional metrics ASAP. More impactful techniques include: • Click Density Analysis (i.e. Heat Maps) • Data Segmentation (to the Nth degree!) • Internal vs. External Search Analysis
- 15. BEHAVIOR ANALYSIS (CONTINUED)1 Data Segmentation Example Data segmentation is the process of taking your data and breaking it down into VERY granular levels. Therefore, you can understand the different nuances and characteristics of your customers. Marketers and Web Analysts use this data to group customers into multiple “segments” and develop increasingly more targeted/relevant marketing campaigns. // The data above is not segmented. The data above is not useful!
- 16. BEHAVIOR ANALYSIS (CONTINUED)1 Data Segmentation Example //The data above is segmented and is very useful! The reports shows that 85% of our search engine traffic is coming from Google. However, 66% of traffic from Google is bouncing! (Bounce defined as a visit < 5 seconds) //As Web Analysts, we should investigate the cause and work to fix this issue. The problem could be that our Google paid search advertisements are underperforming.
- 17. OUTCOMES ANALYSIS2 The goal of outcomes analysis is to answer the question “So What?” “ “
- 18. OUTCOMES ANALYSIS (CONTINUED)2 • For ecommerce websites, Web Analysts measure how many orders were places, what products were purchased, and how much money was made. Equally important, Web Analysts strive to determine why we made as much money as we did. • Conversion rate is the most popular outcome metric, defined as the percent of visitors who completed a success event (i.e. placed an order, registered, completed a form). Kick it Up a Notch Use TRENDED charts with STATISTICAL CONTROL LIMITS when presenting any key performance indicators. You can immediately identify data outliers and deem your campaigns a success or failure.
- 20. EXPERIENCE ANALYSIS3 The goal of experience analysis is to get into the heads of your customers and understand exactly why they do the things they do. “ “
- 21. EXPERIENCE ANALYSIS (CONTINUED)3 • Experience analysis is perhaps the most important pillar of data- driven marketing. It is also the most difficult to implement and the least utilized technique in digital marketing. • Marketing research techniques are used including customer satisfaction surveys, Net Promoter Score, A/B testing, and laboratory- based customer observation. Kick it Up a Notch Measure the TRUE CONVERSION RATE of your website by using a continuous “primary purpose” survey on your website. You will discover that your ecommerce conversion rate IS NOT simply # orders / # visits.
- 23. EXPERIENCE ANALYSIS (CONTINUED)3 Survey Questions A. What task are you looking to accomplish on our website today? (Choose One Radio Button: Research, Purchase, Careers, etc.) B. Were you able to complete your task today? (Choose One Radio Button: Yes, No) C. If you were not able to complete your task today, why not? (Open-Ended Text Box) //This should be a “continuous listening” survey. Use a tool such as ForeSee Results to present the survey to 10% of statistically random website visitors. //ForeSee will also set a cookie to ensure visitors who have completed/declined the survey do not receive another prompt for 90 days.
- 24. EXPERIENCE ANALYSIS (CONTINUED)3 Survey Results
- 27. ADVANCED STRATEGIES @ INTERLINE BRANDS RECENCY, FREQUENCY, MONETARY (RFM) SEGMENTATION A INCREMENTAL REVENUE MEASUREMENT PREDICTIVE ANALYTICSB C
- 28. ADVANCED STRATEGIES @ INTERLINE BRANDS Incremental Revenue Calculation (SAMPLE)