ISDI MDA Master Class November_2015
Download as PPTX, PDF1 like676 views
The document presents a master class on digital analytics fundamentals, covering concepts, methodologies, and key performance indicators (KPIs) essential for optimizing business decisions through data-driven insights. It emphasizes the importance of identifying relevant metrics and segmentation to understand visitor behavior, aligning digital strategies with business objectives, and the critical role of effective communication of results. Ultimately, the document advocates for a structured approach to analytics that fosters consensus and drives meaningful action.
![“The world in [our data] is not a precise
replica of reality (…)”
Paraphrased by me](https://image.slidesharecdn.com/mdamasterclassnov2015-151215010146/85/ISDI-MDA-Master-Class-November_2015-7-320.jpg)
ISDI MDA Master Class November_2015
- 1. Jacques Warren, CEO MDA MASTER CLASS – Madrid 2015-11-19 DIGITAL ANALYTICS FUNDAMENTALS CONCEPTS, METHODLOGIES, KPI
- 2. AGENDA - Positioning Digital Analytics (well, all Analytics) - Some concepts - Practical cases and exercises - Some more concepts - Test - Many more concepts - Making the case for optimization - Then I let you go
- 3. DIGITAL DATA AND ANALYTICS
- 4. Business Without Analytics Business With Analytics Analytics Added Value THE PURPOSE OF ANALYTICS
- 5. Noise Signal Decision Data Insight Action THE PURPOSE OF ANALYTICS Source: Éric Nguyen, Banque
- 6. “The world in our heads is not a precise replica of reality (…)” Daniel Khaneman DATA AND REALITY
- 7. “The world in [our data] is not a precise replica of reality (…)” Paraphrased by me
- 8. What is Analysis? DATA ARE FOOTPRINTS…
- 9. HISTORICAL DATA AND TRENDS HUH!? Outliers are not the enemy
- 10. Visit Attributes Visitor Attributes Customer Attributes DIGITAL ANALYTICS PLANES
- 11. A TOOL PERSISTENT MODEL
- 13. A SIMPLE MODEL INPUT OUTPUTOFFER 2X 2X 1.2X1X ANALYTICS
- 15. A SIMPLE MODEL – DIGITAL VERSUS WHAT? REALM OF CRO REALM OF CRM DIGITAL ANALYTICS DATABASE ANALYTICS DIGITAL CONTENT
- 16. % New Customers % of Subscription Pageviews/Session Units/Orders % Satisfied Clients % Sales to New Visitors % Downloads Subscribers RSS % Email Sales Conversion Rate SEM Ratio AOV/Campaign AOV % Web Sales/Total Sales Sales/Session LOOKING FOR THE RIGHT METRICS
- 17. DEFUSING SOME MYTHS - TRAFFIC – Not THAT important - CONVERSION RATE – Beware of blind optimization - CONTENT IS KING – Nope, processes are - WHAT ARE YOURS?
- 18. A LITTLE BREAK FROM CONCEPTS
- 19. Look at products Shopping Cart Page Order Form ORDERS CASE #1
- 20. CASE #2 IMAGE Bla bla bla bla bla bla bla bla bla bla bla bla bla bla bla bla I WANT IT FORM – Sept 1 Age Destination Trip Worth Email Continue 10% 10% of 10% = 1,2M$
- 21. CASE #3 0 500 1,000 1,500 2,000 2,500 3,000 3,500 1mai,2013 2mai,2013 3mai,2013 4mai,2013 5mai,2013 6mai,2013 7mai,2013 8mai,2013 9mai,2013 10mai,2013 11mai,2013 12mai,2013 13mai,2013 14mai,2013 15mai,2013 16mai,2013 17mai,2013 18mai,2013 19mai,2013 20mai,2013 21mai,2013 22mai,2013 23mai,2013 24mai,2013 25mai,2013 26mai,2013 27mai,2013 28mai,2013 29mai,2013 30mai,2013 31mai,2013 OPENING HOUR PAGE VISITS AND MUSEUM ENTRNCES OPENING HOUR PAGE VISITS MUSEUM ENTRANCES
- 23. SOME DEFINITIONS Measures the evaluate the quality of an organization’s performance in its execution of strategic activities for its present and future success. Applied to Digital, KPIs tell us if the digital strategic vision is executed well.
- 24. KPI CHARACTERISTICS - Align with the online strategy, itself aligned with the business one; - Motivate action; - Allow for prediction; - Be standardized; - Be displayed in context (targets, tolerance threshold, etc.).
- 25. MORE THAN METRICS KPIs are a some kind of language; how we talk about the business. KPIs must be the results of a consensus.
- 26. A PROPOSED METHODLOGY - Reaffirm the digital strategy; - Define and list the expected outputs; - Document and reconfirm consensus; - Validate data quality and availability; KPI WORKSHOP & PROCESS - Decide how results will be communicated.
- 27. KPI & DRIVERS
- 28. KPI & DRIVERS
- 29. EXAMPLES OF THAT TYPE OF WORK DOCUMENT
- 30. EXAMPLES OF THAT TYPE OF WORK DOCUMENT
- 31. EXAMPLES OF THAT TYPE OF WORK WHERE YOU WORK
- 32. AN ANOTHER LITTLE BREAK FROM CONCEPTS
- 33. TEST KWANTYX
- 35. TEST KWANTYX
- 36. TEST KWANTYX
- 37. SEGMENTATION
- 38. ATTEMPTING A DEFINITION - Subdividing a dataset to identify smaller populations with meaningful behavior. - Segmentation is at the heart of the analysis process. It tells us about possible causes of behavior, and where/how to influence them. - Segmentation is to the analyst what dissection is to the physician
- 39. SEGMENTING WHAT? - Segmenting by behavior types or attributes. - Segmentation Levels: traffic, visitors/users, customer file
- 40. LOOKING THROUGH DIEFFERENT ANGLES Basic Digital Analytics Segmentation - Sources
- 41. WHO IS THERE TO DO WHAT? The Importance of Use-Case (Gary Angel’s great contribution) Three principles are the foundations of the two-tiered segmentation approach: Its about understanding the buyer/visitor, not measuring the activity on the website; Intention drives visitor behaviour, so we can reconstruct intention from sequences of actions; Once we have established the visitor’s intention, we can then determine whether they were successful or not in accomplishing their task
- 42. WHO IS THERE TO DO WHAT? We aim to determine WHO the visitor is and WHAT they are trying to do. With these elements, we build use cases. With use cases, it then makes sense to segment KPIs.
- 43. TOWARD THE TWO-TIERED SEGMENTATION Web Site Usage Segments (KWANTYX’s Client) - Information Seeking Prospects - Advanced Prospects – Ready to Convert - Clients Managing Their Account - Clients Adding Services - Job Seekers - Others
- 45. Measuring Marketing is measuring people who do marketing. METRICS ARE POLITICS
- 47. TRUTH CAN YOU HANDLE THE TRUTH?
- 48. THE ILLUSION OF TRUTH
- 50. Source: Gary Angel, EY WELL, TALK ABOUT A PROGRAM!
- 51. Source: Gary Angel, EY WELL, TALK ABOUT A PROGRAM!
- 52. Source: Gary Angel, EY WELL, TALK ABOUT A PROGRAM!
- 53. Source: Gary Angel, EY WELL, TALK ABOUT A PROGRAM!
- 54. Source: Gary Angel, EY WELL, TALK ABOUT A PROGRAM!
- 55. Source: Gary Angel, EY WELL, TALK ABOUT A PROGRAM!
- 56. OPTIMIZATION
- 57. DOES DIGITAL ANALYTICS WORK? Isn’t optimization the whole purpose of Digital Analytics?
- 58. WHAT IS OPTIMIZATION https://www.youtube.com/watch?v=BzLSTpaZkrI Go watch that video. Watch it up to the end, and see what 63 years of optimization could do. Show it to your team next time they say you can’t squeeze any more value. Remember: the 1950 people were at the top of their game…
- 59. Measure everything you do, THE OUTTER LIMITS
- 60. Measure everything you do, but don’t do only what you measure. THE OUTTER LIMITS
Editor's Notes
- #10: -- In behavioral analysis, which is the vast majority of what we do in Digital Analytics, our data is historical. We analyze what happened. This means, among other things, that outliers play a totally different role than, say, in normal statistical analysis. -- Sure, outliers can be accidents, values out of the norm that need to be discarded, but they can very well be (and often are) interesting manifestations of unplanned, unpredictable things. In Digital Analytics, we can’t ignore outliers, we can’t risk NOT to discover unexpected phenomena which could have value.
- #34: CORRELATION: 0.85