How to Come Up with Winning A/B Tests Using Data

One way to increase your sales is to improve your conversion rate, right? Although it’s true, conversion rate optimization isn’t easy.

If you just base your tests on your gut feelings, you may find a few winning variations, but chances are, most of them will lose.

Download a free print ready version of this infographic.

So, how you do improve your conversions? You have to analyze data before you run A/B tests. In order to help you improve your conversion rate, I decided to enlist the help of Peep from ConversionXL and his brilliant ResearchXL framework to come up with an infographic that breaks down what you need to do.

Click on the image below to see a larger view:

How to Come Up With Winning A/B Tests Using Data

Click here to view an enlarged version of this infographic.

Conclusion

I run hundreds of A/B tests, and the variations that typically have the most impact are those with drastic changes. To come up with those drastic changes, you need to analyze your data first to see what you need to change.

You can get the data by looking at your Google Analytics reports as well as a heatmap of your site through Crazy Egg. In addition, consider running surveys by using tools such as Qualaroo.

Have you run any A/B tests yet?

PS: Make sure you checkout this framework.

Embed This Image On Your Site (copy code below):

If you want to break through to real profits online, you need some serious firepower. For a limited time I’m sharing some select tips and tricks Amazon, Microsoft, NBC & Hewlett Packard paid thousands of dollars per hour for, FREE.
  • The step by step guide to monster traffic generation
  • The how-to guide for increasing conversions on your website
  • 7 Cashflow killers your analytics tools are hiding from you
     
 
100% privacy, I will never spam you!

Comments

  1. Eduardo Cornejo :

    Hey Neil,

    Awesome post. I’ve been using ThriveLeads for opt-ins, and will soon start A/B testing with it.

    Also, how much traffic would I need to do a useful survey?

    Thanks!

    • Eduardo, It really depends. I’ve run surveys with small data sets and of course others with large data sets. You can run surveys at any point but make sure the numbers and answers are all in context so you can get the best analysis from your results.

      • Another great infographic.

        I had noticed that all our clients have different LP requirements but only A/B testing can increase conversion rates.

        • Anil Agarwal :

          Conducting A/B split tests is the surefire way to increase conversion rates of any website.

          Last time, when I used GetResponse to perform A/B split testing, I saw amazing email open rates with two variations. Will definitely give another try for doing A/B tests on my website landing pages too.

  2. Steve Estimable :

    Neil, Insightful infographic on winning ab tests using data.

    I like the comparison you made on Wrong Mentality vs Right Mentality.

    Neil Q: Have you run any A/B tests yet?

    Steve Q: Yes, lately more on social media.

    Have a great Friday!

  3. Randy Kauffman :

    Very Nice! I love having processes. They are easy to follow and you can train anyone to use them.

    Thanks Neil!!

  4. William Zimmerman :

    Neil,

    Great post man!! Again!

    Love the mouse tracking analytics. Will be sure to use Crazyegg when my website it up and running.

    Have an awesome weekend!!

    All the Best,
    Bill Zimmerman

    • Bill, glad you found the post helpful. Let me know how it works out — I look forward to hearing much more from you on results.

  5. Reese Garcia :

    Neil,

    Thanks for laying out a clear process for A/B testing. As a PPCer, I appreciate the formulas you included to help verify whether an adjustment is worth its return.

    The section on speed testing reminded me of a case I read where Amazon had calculated that every .1 seconds of loading time was worth $1000 in revenue to them. Those might not be the exact numbers, but the story shows how relevant speed can be to UX (so long as the right foundation is in place.)

    Thanks for the great post, Neil.

    • Reese, glad you found the information helpful. Speed is definitely a factor that needs to have greater weight placed on it. Thanks for the feedback and insights. If you come across any other nuggets of insights please let me know.

  6. Palash Kumar Daw :

    Sorry i didn’t run any A/B test 🙁

  7. Anshul Raghav :

    Amazing article!! Now I will apply this technique on my landing page of my Technology, Guide Blog.

    Thanks

    • Anshul, glad you found it helpful. Please let me know if you need help with anything else. Looking forward to hearing much more from you.

  8. Ahfaz Ahmed :

    Hi Neil,

    Great Infographic. This will surely help a lot in testing and making things work much better. I never heard about mouse tracking. Will definitely use Crazy Egg to analyse my website.

    • Ahfaz, — great. Let me know how it works out and if you need help along the way. Looking forward to hearing more from you.

  9. Adarsh Sojitra :

    The best piece od art as well as the informative piece I have ever seen on Quicksprout. This infographic is Just a great! with lots of information.
    Thanks neil for this….

    • Adarsh, glad you found it helpful. If you have any questions or need help with anything please let me know.

  10. Rahul Sinha :

    I am ux designer even I didn’t know all these stuff, thanks for sharing such a valuable information with us, I liked the right and wrong mentality 🙂

    • Rahul, glad I could help. Looking forward to hearing much more from you. As always if you need help with anything please let me know.

  11. Alecia Stringer :

    I love the formula and resources shared here. Thanks.

  12. Marston Gould :

    While I am a big fan of testing, one thing that I rarely see anyone talk about is how not to be fooled by the results of a test. All too often I hear people say that they ran a test, one version clearly won, but then after a while, our results faded. While you would think that they would understand what variable is coming in to play, most people miss the obvious – which is – change itself. Sometimes, it does not matter what your ‘test’ is. If it is well conceived – at least as well conceived as your control, it will often win, for a short time.
    This is why I always suggest to people that they build their testing process around trying to understand the value of change. There is a cost to constantly pushing or keeping performance high. Is that cost worth the performance that could have been had by a solid winner performing for over the long haul? Or perhaps you just put winners in rotation.

    Either way…nice infographic

  13. Alex Schulko :

    Neil,

    This is fantastic! I am currently assisting our internal CRO team at Perficient, Inc. out of Milwaukee to improve conversions for our specific clients. This is a new team and department, so we are still learning as we go.

    I immediately shared this infographic with the team as I’m certain this information will help us determine the right elements, funnels, and tasks to improve on our clients websites.

    Thanks again so very much for all of your insightful knowledge Neil!

    Kind regards from a loyal follower,

    Alex

  14. Amazing article !! I really love the formula and resources shared here !!

  15. Webmastering Badass Eric G. :

    Hey Neil this is great is it alright to embed your infographic within my page? I will surely link back to the url! Keep up the great work. Eric

  16. Hi Neil,

    Very interesting post again.
    I really like the way you present the information in the infographic. It’s packed with value and also easy to consume an digest this way.

    Thanks a lot,
    Elmar

  17. Great post! I haven’t performed any A/B test yet, I’m going to look deeper into this area.

    Aside from all the benefits of infographics, natually when you had one, the number of total words are reduced. I guess the benefits of having one outweighs the impact of reduced words. Is it better off to still have some paragraphs but still have the infographic to summarize?

    Thank you!

    • Robin — great question. I have done testing on that question and have found that it’ sbest to have paragraphs in my case.

  18. Harshit Bhootra :

    Hey neil !
    a good Infographic it was 🙂
    I too started my own WordPress blog ^_^
    learning all the way by you 🙂

  19. Scott Boren :

    Neil,

    Love all the data driven posts here and on http://www.neilpatel.com, you always seem to get at the root of an issue in a concise and practical way.

    I spent my pre-entrepreneur years in insurance product management … talk about big data!

    I was wondering, so many who may find your blog useful … like myself … may (and I stress may because I don’t know) be less analytic. Do you find that bloggers and other small business owners struggle with the use of analytics and data?

    • Scott — I definitely think it’s a tough balancing act to use analytics and balance that with content creation. Please let me know if you need any help along the way.

  20. Bilal Tahir khan :

    Hi Neil,
    very interesting infographic.very helpful for me nice and easy formulas.you guide very well.
    Thanks for guide and sharing keep up it.

  21. Hi I’ve been a reader of your blogs and I think they are the best information about. Although I would like you to explain what the code is that you ask readers to embed on their websites please? I don’t know what it is. Thank you.

  22. keith patterson :

    Neil, you seem line an honorable guy, so I assume that whomever you had write this post did not tell you that they ripped it off Conversion XL, literally word for word, and dressed it up with graphics.

  23. Hi Neil,

    Thanks for coming up with a blog on A/B Testing as this makes a great intro on the topic for me. The bulk of my marketing experience lies on performing market surveys in the traditional sense as it is done on exporting.

    No time to lose, I got a lot to learn to help our exporters set-up a phenomenal eCommerce one-stop shop.

    thanks heaps!

    Gigi

  24. harekrishna :

    The advantages of A/B testing are truly helpful to improve conversions. Roles of Google Analytics, which you have mention including technical analysis and speed analysis are very significant. I am planning to focus on A/B testing.

  25. A very nice Infographic, which explains the topic in right detail. I specially liked the Mouse tracking analysis as didn’t know about it earlier.Also the right mentality & wrong mentality section was insightful too.I gained a lot of knowledge from them.

    Best wishes for your next post!

  26. Hello Neil. My name is Dan and I am a founder of Self Employment Ideas website. Many thanks for the great A/B Split Testing article and lovely Infographic. It’s a great value and a perfect example for the people looking for the fresh ideas, the new on-line marketing techniques and proper guidelines. I think it’s inspiring and revealing and I will add it to my bookmark list.
    Keep up the good work. Thanks and do have a great day!
    Regards,
    Dan
    Founder of http://www.selfemploymentideas365.com

    • Dan, glad you found the post helpful. Please let me know if you need help with anything at all.

  27. Pardeep Goyal :

    Once again Epic post.

    I have not thought of doing A/B test on my blog because I have very limited audience.

    How much audience is good to have before even thinking about A/B test?

  28. I have never thought about A/B Testing.. But now I understand.
    Thanks for the info graphic

  29. Sorry i didn’t run any A/B test. i am learning lots of things from every article.
    Thanks for sharing nice content.

  30. Wow, very detailed post! Does publishing in the form of an infographic hurt your SEO, since it’s basically one big image, rather than a bunch of text that Google can read? How does that work?

    Thanks for all the great posts!

    Matt

  31. Prabhu Tharmarajan :

    Great article! I’ve read about A/B testing, and how it can generate leads for companies. The important thing is to not stick to one test, but multiple tests, and come up with a winning test based on the data gathered from them. A few tweaks here and there, and you’ll have one that will give you a definite winner!
    Thanks for the article, it really is enlightening!

    • Prabhu, great point — the ultimate goal is to use as many tests as possible to get the best results.

  32. Hi Neil,

    I was unable to control myself from commenting. I must say your info graphics are just amazing, it provides to much of value to the post. Moreover these info graphics makes your post looks interesting and appealing. Just read your mail regarding this post, the words where appealing, although the format remained same. But, I have to say there is lot to learn from you. Best of luck 😀

    MS Qureshi

  33. Thanks for this cool infograph, I have been searching this for a long time now.

  34. It’s really great information on How to Come Up with Winning A/B Tests Using Data. I like the presentation above. It’s very inspiring to read this kind of blog. Thanks for sharing. God Bless.

  35. maha 10th results 2015 :

    Do you recommend any tool for A/B testing neil?

  36. A very nice Infographic, which explains the topic in right detail. I specially liked the Mouse tracking analysis as didn’t know about it earlier.Also the right mentality & wrong mentality section was insightful too.I gained a lot of knowledge from them.

  37. This is a great post. This post give truly quality information. I am definitely going to look into it. Really very useful tips are provided here. Thank you so much for sharing and keep it up the good work.

  38. Hi,

    Thanks for this illustrative infographic. A/B testing is a great way to test the best positions and layouts.

  39. Neil,great post. I just wanted to ask you one question. In the last posts you mentioned that first one should concentrate on traffic generation rather than focussing on revenue. So,my question is at what instant shall i change my priority to revenues. Eg. 5000 visitors a day or 10000 visitors a day or maybe 1000/2000 subscribers? Please help me .

  40. thanks for sharing

  41. I Just want to ask you a queation that how much time do you require to write such a long blog post including content research, image or infographic creation, content creation etc.

  42. Hi Neil – Fan of your work man. Its generally really good information. I am having an issue with a landing page for my biggest client. I can see in analytics that the conversions for a particular page are far below what would be expected on the rest of the site.

    Unfortunately im trying to optimise the page based on keyword intent rather than on behaviour, this is due to lack of resources available and the old system the site is built with.

    Instead im going to have to implement the best landing page I can, then replace in around 1 month, and split test that way.

  43. Hello neil sir, can you tell me how to rank event blog. my blog is: http://www.happyvalentineweek2017.com/

    • You need to connect with your audience and influencers in your industry 🙂

      • I am really a busy schedule with them, But don’t have any time table, so please sir i need your help, Because as i know you are managing lot’s of website at once.So please some tips, Some Guru Mantra PLZZZ.

  44. Hi Neil Sir, May I get some suggestions from you on my Education Blog https://www.gkjobs.in/ how to rank it, though I written many articles !

Speak Your Mind

*