Engage 2013 - Give Your Optimization Program Wings
0 likes465 views
This document discusses how to optimize a travel website. It recommends identifying goals upfront and beginning with brainstorming test ideas. An example is provided of how testing small changes to button text and page layout on the booking confirmation page led to booking lifts of 2.6-8.69%. The document advocates mixing simple and complex tests across markets, journeys, and devices and revisiting priorities quarterly based on new data. An example EU cookie law test is described which found an optimal notice message and placement.
![4. EXAMPLE: Mobile & Tablet – make a case
+61.32%
“Despite the [rapid Search
growth of mobile] , 69
companies in the FTSE
100 had done nothing
to “optimise” their
websites for mobiles…”
37](https://image.slidesharecdn.com/engage-2013give-your-optimization-program-wings-130205181224-phpapp02/85/Engage-2013-Give-Your-Optimization-Program-Wings-37-320.jpg)
Engage 2013 - Give Your Optimization Program Wings
- 1. How to give your Optimisation Program Wings Frans Poldervaart Tom Waterfall Web Analytics Manager Director of Optimisation Solutions
- 3. - 75 country websites - 23 different languages - 1100 pages per country - Largest travel site in 175.000 Netherlands (Twinkle T100) transac-ons per 24% online month KLM.com: By the numbers 3
- 4. 4
- 5. 1. How to get started PEOPLE DRIVE SUCCESS
- 6. 1. How to get started: Identify Your Goals Upfront • Increase share in online sales • Use the most effective acquisition channel • Increase flying blue members • Make it easier for customers to navigate and purchase • Increase ancillary revenue • One stop shop • Ensure customers’ experiences are more positive 6 Name of presentation - Version 1.01 dd-mm-2008
- 7. 1. How to get started 7 Name of presentation - Version 1.01 dd-mm-2008
- 8. 1. How to get started
- 9. 1. EXAMPLE: How we took off 1 Control Functional 2 Inspiration Trust 3 4 9
- 10. 1. EXAMPLE: How we took off Inspiration +14.11% +5,400 +€3.8 Conv. Bookings Million 10
- 11. 1. How to get started: Following Take Off • Begin Brainstorming test ideas • Use all the tools at your disposal 11
- 12. 12 Name of presentation - Version 1.01 dd-mm-2008
- 13. 2. Getting buy-in: Start an illegal gambling ring 13
- 14. 1. EXAMPLE: Starting an illegal gambling ring 14
- 15. 2. Getting buy-in: Infiltrate & Cultivate 15
- 16. 2. Getting buy-in: Shout about results & learning 16
- 17. 2. EXAMPLE: Shout about results & learning 17
- 18. 2. EXAMPLE: Shout about results & learning +23% Lift 18
- 19. © 2011 Webtrends, All Rights Reserved. | 19
- 20. 3. Planning & Prioritization: Companies struggle 20
- 21. 3. Planning: Focus on quick wins This chart is based on tentative information presented, and elements might move around based on new data available. Starbursts indicate Booking Details page tests HIGH revenue lift potential. HIGH Booking Details page tests Simple Complex Changes to structure, arrangement or MID functionality Search results page functionality Tests with filters, modifying search MID Strikethrough pricing Affecting display, but only doable during sales Exceptions testing Search results page display ? No results pages etc. MID Promo codes, supplier logos, messaging Across the site ? Inclusions, messaging of any kind Homepage tests Any display/messaging HIGH Related tests Revenue Potential 21
- 22. 3. EXAMPLE: ROI Potential vs. Complexity 22
- 23. 3. Planning: Prioritise a mix of simple & complex 23
- 24. 3. EXAMPLE: Mixing Simple and Complex +2.6% Lift
- 25. 3. EXAMPLE: Mixing Simple and Complex +39% Lift
- 27. 3. EXAMPLE: Mix & Match Control 27
- 28. 3. EXAMPLE: Mixing Simple and Complex Level 1 Level 2 Level 3 Level 4 (Control) Lift / Booking ‘Next step’ Lift / Level Chance to Beat Control Confirmation conversion rate Chance to Beat Control Rate 1 33.23% --- 3.41% --- All alternate button copies drove significantly more visitors to the next +5.05% 2 step 34.91%also had a chance to beat impact on Bookings. and (99.43% positive 3.21% -5.84% (Not Significant) control) +5.48% +8.64% 3 35.05% (99.64% chance to beat 3.71% (85%% chance to beat control) control) +8.69% +6.04% 4 36.12% (99.79% chance to beat 3.62% (Not Significant chance to beat control) control) 28
- 29. 3. EXAMPLE: Mixing Simple and Complex Level 1 (Control) Level 2 Underlining City and Price and making the whole row clickable Lift / is driving significantly more visitors to theConfirmation Level ‘Next step’ Chance to Beat next step,Chance tois Booking and / Beat Lift conversion rate Control showing promise with Bookings Rate Control 1 33.76% --- 3.37% --- 6.47% +6.69% 2 35.95% (99.99% chance to beat 3.60% (87% chance to beat control) control) 29
- 30. 3. EXAMPLE: Mixing Simple and Complex +150 Bookings +6.69% Lift The ‘Clickable Row’ had the highest impact at the booking confirmation rate, with a +6.69% lift 30
- 31. 3. Planning: Mix & Match between markets/journeys US Traffic Rest of Bookings World Car Hire English Spanish Loyalty Registration Email Sign-Up Mobile / Tablet Site 31
- 32. 3. Planning: Revisit priorities quarterly 32
- 33. © 2011 Webtrends, All Rights Reserved. | 33
- 34. 4. EXAMPLE: EU Cookie Law Test Background 34
- 35. 4. EXAMPLE: EU Cookie Law Test Background Control Level 4 Optimal Experiment Conversion Confirmation page Conversion Run Date Click button Page Lift (confidence interval) Lift Control - --- 2.98% (+/- 0.21%) --- 12/11/2012 – 26/11/2012 Optimal 2.43% (+/- 0.07%) - 3.09% (+/- 0.21%) +3.64% 35
- 36. 4. EXAMPLE: Mobile & Tablet “My unborn nephew owns a smartphone,” says person on street. 36
- 37. 4. EXAMPLE: Mobile & Tablet – make a case +61.32% “Despite the [rapid Search growth of mobile] , 69 companies in the FTSE 100 had done nothing to “optimise” their websites for mobiles…” 37
- 38. 4. EXAMPLE: Mobile & Tablet – fix it 38
- 39. 4. Be Thoughtful: Track more, Learn more 39
- 40. 4. EXAMPLE: Track more, Learn More When the “Filter by Location” option was amended to a dropdown… …visitors used the ‘Star Rating’, ‘Filter by Board’, and Map tools significantly more. Explanation: the dropdown takes up far less real estate and therefore places further emphasis on other filters. Additional Insight: the increased usage of star rating, by board and the map tool correlated positively with an increase in hotels booked. Recommendation: continue to test filter ordering, emphasis, and style to identify the real drivers behind hotel bookings. 40
- 41. 5. The Path Ahead for • From 0.5 FTE to 2 FTE dedicated to the conversion optimization program • Large and complex testing continuing with full support from experts • Small tests conceived and built by KLM • Increase communication within the organization • Proactive testing associated with site developments • Responsive design • New applications 41
- 42. Lessons Learned • Takeoff requires a central owner and initial testing that focuses on business value • Attain departmental buy-in through a healthy mix of sharing and gambling • Plan ahead, and revisit your plans regularly • Keep it simple and look at all your journeys, markets and goals • Keep your program fresh through innovation and relevance • Once you’ve taken off, don’t plan on ever landing!
- 45. © 2011 Webtrends, All Rights Reserved. | 45