UX Analytics and Experimentation for eCommerce Growth
Download as PPTX, PDF0 likes989 views
The document discusses the importance of UX analytics and experimentation in driving ecommerce growth, highlighting the experience of Narayan Keshavan in managing digital analytics for Dell. It emphasizes the significance of iterative experimentation and understanding user feedback, warning against 'testing myopia' in the analysis of test results. The key takeaway is that every test, regardless of outcome, provides valuable insights that can enhance the user experience and drive conversion rates.
![[Next up]
An Interview with Amazon Prime Video’s
Sr. Product Lead Karan Peri
Date & Time - 22 July | 11 AM ET / 8 AM PT
Details - vwo.com/webcast](https://image.slidesharecdn.com/uxanalyticsandexperimentation-narayankeshavan8july2020-200708160451/85/UX-Analytics-and-Experimentation-for-eCommerce-Growth-37-320.jpg)
UX Analytics and Experimentation for eCommerce Growth
- 1. UX Analytics and Experimentation for eCommerce Growth Narayan Keshavan Head – Digital Analytics (Design Analytics)
- 2. About Me • Head of Digital Analytics in the Design Analytics org., a global team that owns the experimentation charter for Dell.com • Over 13 years of experience spanning eCommerce, Telecom and Insurance sectors • Have managed marketing with P&L responsibilities for over USD 45 Mn in the hyper-competitive telecom sector • Have managed teams working in the area of marketing and digital marketing analytics including conversion rate optimisation, campaign analytics, SEO and social media listening. • Lead the Digital Analytics Community (DAC) initiative to build and expand a thriving community of digital analytics professionals DigitalAnalyticsCommunity.wordpress.com
- 3. Focus on the user EDD EDDand all else will follow
- 4. Dell.com - an optimization journey with experimentation at its core 1996 2009 2020 Narayan Keshavan | Design Analytics | Dell Digital Design
- 5. A “great idea” can come from anywhere Narayan Keshavan | Design Analytics | Dell Digital Design Quantitative Qualitative UX Analytics Heatmaps/ Session replays VOC Test iterations Business inputs
- 6. Building your experimentation roadmap Experimentation Roadmap Business inputs Bench- marking Voice of users Product roadmap Test Insights Analytics UX Research Product managers Design/ UX Marketing eCommerce Segment/ Element targeted KPI Projections
- 7. Iterative experimentation success metricshypothesis analysis performance check
- 8. Key principles for an effective experimentation strategy
- 9. Less is ‘More’? Please note: the screenshots have been simplified for illustration and do not completely resemble the actual version
- 10. Nothing is OBVIOUS Experimentation is the real ‘Moment of Truth’ and can, many times, SURPRISE us! Image courtesy: Fabian Reitmeier | pexels
- 11. Reducing the clutter had unintended consequences Please note: the screenshots have been simplified for illustration and do not completely resemble the actual version
- 12. “I have not failed. I've just found 10,000 ways that won't work.” ― Thomas A. Edison
- 13. There is no such thing as a ‘failed experiment’! Every test offers insights which feed into subsequent iterations
- 14. We tried to simplify the checkout pages by collapsing optional fields optional fields Please note: the screenshots have been simplified for illustration and do not completely resemble the actual version
- 15. Interestingly, more users dropped off later in checkout optional fields collapsed Please note: the screenshots have been simplified for illustration and do not completely resemble the actual version
- 16. The insights gained helped us iterate with a better experience optional fields at the bottom (not collapsed) Please note: the screenshots have been simplified for illustration and do not completely resemble the actual version mandator y fields
- 17. Beware of ‘Testing Myopia’Avoid ‘Testing Myopia’ Short-sightedness in interpreting test results may prove detrimental to business Image courtesy: David Travis | unsplash
- 18. Let’s consider an A/B test for a banner Variation A Variation B
- 19. Which variation is better? Do you say ‘B’? A B Metric Variation A Variation B Visits 1000 1000 Users 200 100 Conversion Rate Users 10% 12% Please Note: I have taken random numbers for convenience of explanation. These are not from any real test
- 20. Which variation is better? Is it that obvious? A B Metric Variation A Variation B Visits 1000 1000 Users 200 100 Orders Users 20 12 Conversion Rate Users 10% 12% Please Note: I have taken random numbers for convenience of explanation. These are not from any real test
- 21. What do you think now? Is ‘A’ better? A B Metric Variation A Variation B Visits 1000 1000 Users 200 100 Non-users 800 900 Orders 36 30 Users 20 12 Non-users 16 18 Conversion Rate 3.6% 3.0% Users 10% 12% Non-users 2% 2% Please Note: I have taken random numbers for convenience of explanation. These are not from any real test
- 22. What do you think now? Is ‘A’ better? A B Metric Variation A Variation B Visits 1000 1000 Users 200 100 Non-users 800 900 Orders 36 30 Users 20 12 Non-users 16 18 Conversion Rate 3.6% 3.0% Users 10% 12% Non-users 2% 2% Please Note: I have taken random numbers for convenience of explanation. These are not from any real test
- 23. Users are on your team Image courtesy: Malte Wingen | unsplash Keeping your ears open to what users are saying is often rewarding
- 24. Users were not sure about the payment options available till they reached the payment page Please note: the screenshots have been simplified for illustration and do not completely resemble the actual version
- 25. Simply adding this information on the cart page improved cart conversion Please note: the screenshots have been simplified for illustration and do not completely resemble the actual version
- 26. Competition has it; do we need it too? Please note: the screenshots have been simplified for illustration and do not completely resemble the actual version
- 27. Every website is unique ‘Competitive Benchmarking’ trap - just because a leading website has it, doesn’t mean it will work for you
- 28. We added merchandising banners in the masthead
- 29. Only to realise that not many users clicked on it
- 30. Don’t lean towards ‘your’ idea Aim to ‘TEST' if the idea is effective; not ‘PROVE' that the idea is effective Image courtesy: Marco Ceschi | unsplash
- 31. What is the best way to sort search results? Please note: the screenshots have been simplified for illustration and do not completely resemble the actual version
- 32. Conversion improved but not for the reasons expected Please note: the screenshots have been simplified for illustration and do not completely resemble the actual version
- 33. Did we implement this then? Please note: the screenshots have been simplified for illustration and do not completely resemble the actual version
- 34. Every website is unique Don’t lean towards ‘your’ idea Nothing is obvious There is no such thing as a ‘failed experiment’ Users are on your team In a nutshell… Avoid ‘Testing Myopia’
- 35. Thank you Thank you
- 36. QnA
- 37. [Next up] An Interview with Amazon Prime Video’s Sr. Product Lead Karan Peri Date & Time - 22 July | 11 AM ET / 8 AM PT Details - vwo.com/webcast