Se ha denunciado esta presentación.
Utilizamos tu perfil de LinkedIn y tus datos de actividad para personalizar los anuncios y mostrarte publicidad más relevante. Puedes cambiar tus preferencias de publicidad en cualquier momento.

How to Create Better A/B Tests and Why You Shouldn’t Bank on Best Practices

2.203 visualizaciones

Publicado el

The web is full of best practices and results of AB-tests. Can’t you just copy those, lie back and watch your conversion rates go up? Bad idea. Best practices don’t always work.
So, what to do instead? Just test every idea that pops into your head? No again! That will result in bitter disappointment because your tests will fail over and over again.

The key is to get to know your visitors before you start testing. Research their behavior on your website. Learn how to create AB-tests that really make a difference thanks to user research. So you can focus on what your visitors want: their needs, their priorities and their actions.

Publicado en: Internet
  • Inicia sesión para ver los comentarios

How to Create Better A/B Tests and Why You Shouldn’t Bank on Best Practices

  1. 1. How to Create Better A/B Tests & Why You Shouldn’t Bank on Best Practices Karl Gilis @agconsult
  2. 2. Some of our clients
  3. 3. Part 1 The joy of A/B testing? @agconsult
  4. 4. It's not because you own a typewriter, that you'll become the next best-selling author.
  5. 5. Welcome to my world
  6. 6. The world of AB-testing
  7. 7. 1. Don't test every idea that pops into your head
  8. 8. Because most of these ideas are stupid
  9. 9. 2. Don't copy tests you've seen on other websites
  10. 10. Tell users what (you want them) to do One of the oldest tricks in the book… @agconsult
  11. 11. 72,3% more clicks to product pages
  12. 12. 13,1% more clicks to product pages
  13. 13. Yes. I have found my holy grail of category page optimization. I now will become world famous and people will ask me to reveal my secret everywhere.
  14. 14. Nothing. Zilch. Nada.
  15. 15. 3,2% drop in conversion
  16. 16. Put a reassuring message near your call-to-action Make sure you use positive words. You don't want to frighten the visitor. @agconsult
  17. 17. 23,2% more conversion 19,4% more conversion
  18. 18. This is an A/B test from Michael Aagaard. Find more details at
  19. 19. 3. Don't blindly follow so-called best practices
  20. 20. Green check marks always work! Really? @agconsult
  21. 21. No matter what the bullets said, the green checkmark versions always won.
  22. 22. And so the green checkmark myth was born  This product sucks  Burning your money is better than buying this sh*t  Read my lips: DO NOT BUY THIS! @agconsult
  23. 23. Nothing. Zilch. Nada. -5,1% -5,9%
  24. 24. Nothing. Zilch. Nada.
  25. 25. "AB-testing isn't a research tool. Use it to validate your homework." Karl Gilis - @AGConsult
  26. 26. "The more I learn, the more I realize how much I don't know." Albert Einstein
  27. 27. Be careful with graphs and numbers like these. There is no such thing as the ideal number of fields.
  28. 28. You have to find out what works best for your website @agconsult
  29. 29. You have to find out what works best for your audience @agconsult
  30. 30. You have to find out what works best for your product or service @agconsult
  31. 31. Do your user research
  32. 32. Part 2 Solid user research = better AB-tests @agconsult
  33. 33. My preferred methods & tools • User testing (Morae) • Eyetracking (MyGaze eyetracker, Tobii) • User Analytics (Google Analytics) • User session recordings (Mouseflow, Inspectlet, Hotjar) • Mouse heatmaps (Hotjar, CrazyEgg) • Form analysis (Formisimo, Hotjar) • Targeted mini surveys (Qualaroo, Hotjar,…) @agconsult
  34. 34. User testing with eyetracking
  35. 35. + 45,9% clicks on "Get a quote" + 15,6% clicks on "Book a test drive"
  36. 36. Put your calls-to-action where people see them I know this doesn't sound very innovative. But it’s really important. @agconsult
  37. 37. • 83% more clicks on 'Get a quote' • 44% more on 'Where to buy'
  38. 38. The right-hand bar is probably not the best place to put your main CTA @agconsult
  39. 39. Google Analytics
  40. 40. 78% more clicks to a product page
  41. 41. Use bullet lists Especially on category pages and above the fold on product and landing pages. @agconsult
  42. 42. + 10% brochure requests
  43. 43. Don't hide (important) stuff on mobile People do scroll. Don't worry. @agconsult
  44. 44. 25% of all clicks 20% 35% 2 to 5%
  45. 45. 28% more clicks
  46. 46. Sliders suck. Always. Except when they are just image slides. Without separate messages. @agconsult
  47. 47. 55% more clicks in the same screen real estate, after reducing the number of messages from 4 to 2.
  48. 48. Don't overwhelm your visitors Less is more. White space has never killed a visitor. @agconsult
  49. 49. "Less is more. White space has never killed a visitor." Karl Gilis - @agconsult
  50. 50. On this landing page, we deleted the floating main navigation. This resulted in an uplift of over 24,2%.
  51. 51. Removing the main navigation for AdWords visitors, resulted in a significant lift in bookings. (Sorry, I'm not allowed to share the exact number.)
  52. 52. Remove navigation for people coming from specific search queries When navigation turns into a distraction, remove or hide it. @agconsult
  53. 53. "Web design isn't about adding stuff. It's about keeping only those elements that add to the bottom line." Karl Gilis - @agconsult
  54. 54. User session recordings & mouse heatmaps
  55. 55. Name of the tool: Hotjar
  56. 56. Only 52,41% of visitors scrolled to this part
  57. 57. Original
  58. 58. Variation
  59. 59. 32,5% more people reach this point: - Our version: 69,45% - Original: 52,41% 44,1% more people reach this point 100% more visitors go to a product detail page 76,2% more requests for quote 57,8% more brochure downloads
  60. 60. Use tools to find where the problem areas are And use your brain or user input to make them better. @agconsult
  61. 61. 43,8% more clicks
  62. 62. Repeat your CTA at the bottom of the page Believe me. This works. Almost always. @agconsult
  63. 63. Form analysis (& user session recording)
  64. 64. Formisimo reveals which fields caused most problems
  65. 65. 52,9% more requests for offer
  66. 66. Make forms as easy as possible Use tools to find out where the real problems are. Don't guess. @agconsult
  67. 67. When we made those fields optional there was a decrease with 6,3%.
  68. 68. And now WHAT?
  69. 69. 25,4% more submitted requests for proposal
  70. 70. Cool, we'll fix that in June 2016
  71. 71. 14,2% more submitted requests for proposal
  72. 72. Keep on testing. Iterate! Your first variation will almost never be the best one. @agconsult
  73. 73. Targeted surveys
  74. 74. Some of my favorite questions • On website • What is the purpose of your visit to our website today? • What are the most important things when buying …? • What convinced you the most to decide to ask a quote? Was there something that made you hesitate? If so, what? • Existing clients • What would you miss the most when your X broke down / when we took away X • What do you like the most about X • What's the most important thing X helped you to achieve?
  75. 75. On the homepage of a webshop, do you push products or product categories?
  76. 76. What is the purpose of your visit to our website today?
  77. 77. + 188% more clicks + 75% more sales from homepage visitors
  78. 78. Push product categories instead of individual products This depends from site to site. But most of the time it works. Big time. . @agconsult
  79. 79. What is the purpose of your visit to our website today?
  80. 80. 23 x more people use the dealer locator
  81. 81. Give people what they want As long as it helps your business. . @agconsult
  82. 82. 14,3% more dealer searches
  83. 83. Use the same words your users do Not the ones you use. . @agconsult
  84. 84. What convinced you the most to decide to ask a quote? Was there something that made you hesitate? If so, what? 77,1% more requests for quote
  85. 85. Know what your clients & visitors want And get filthy rich  . @agconsult
  86. 86. “Stop selling your products/services the way you want to sell them. Sell them the way people want to buy them.” Karl Gilis - @agconsult
  87. 87. Find it!
  88. 88. Get my free ebook now  10 conversion tips that work on every website  15 cases from the real world  6 best practices from leading websites   Already more than 3.216 downloads
  89. 89. That's it! Thx @agconsult