150+ A/B email split testing ideas you can use today

150+ A/B email split testing ideas you can use today

Have you ever received an email campaign and you thought “Wow, this is really great!” ?

Maybe it had a nice design, maybe a strong story, maybe it was very personal.

Let me ask you another question, before I hop on to giving over a 150 email split test examples.

Did you buy from that lovely mail?

As marketers, we know that a single on the mark email can outperform your average newsletter by a mile.
But it isn’t a beauty contest to see who can make the best-looking email or even the smartest email.

The challenge is to get the most engaging result rich email possible.

You might be thinking, simply said, but Jordie which email will do better than others?
Should I do this version or that version?

The ultimate answer is simple: You come up with one great idea.
and then (this is the trick) with a second great idea. The second idea can be a different idea or a variation of idea 1.

Then you let those (and possibly version 3,4,5 and 6, battle it out for the win by running a split test.

Can you lift all your email campaign results to that top level? That all depends on the quality of your ideas.

Know what works with A/B email split testing

Knowing what works is key to insightful marketing campaigns. With A/B split testing you have a great result booster at your disposal.

A/B testing will help you pinpoint which version of your email campaigns is the most effective.

But marketers often get stuck or run out of fresh ideas for email split testing, making it repetitive. So I often get asked for examples of good email split tests. Econsultancy research suggests that there are only a few email split tests that are often done by marketers… which is a shame seeing there is so much more to explore. A good reason to try the fresh email split testing ideas below!

150+ A/B email split test ideas

From the subject line all the way through the spectrum of design, copy, offers. Of course the whole anatomy of the email can be tested. Even the core message and your segmentation.
Below is a list with 150+ ideas for A/B email split tests to try. Some I used before with clients and come from own experience, use it for inspiration.

  1. Product image variations
  2. Soft sell (benefits) versus hard sell (buy now)
  3. Pre-header text
  4. Style based versus Product based
  5. Headline copy and length
  6. Tell the CEO that he cannot write a lame intro this time
  7. Header height
  8. Pricing and discount variations (10% off versus 10 dollars off)
  9. Optimize for Mobile readers
  10. Responsive versus non-responsive design
  11. Product image sizes
  12. One column vs two vs three column
  13. A negative pitch (loss aversion)
  14. The order of your links and CTA’s
  15. Adjusting the colour scheme
  16. Using (price) ribbons
  17. Add odd-shaped arrows and writing to highlight important parts
  18. Incentives to stimulate engagement
  19. Do a radical redesign
  20. Send time of day
  21. Use price brackets
  22. Use sorting cues as categories
  23. Reword your body copy
  24. Upsell based on previous products purchased
  25. Change product description copy
  26. Offer a trial or test membership / purchase
  27. Balance of content in email versus landing page
  28. Feature one product versus multiple products
  29. Animation – animated gifs
  30. Isolate one feature or benefit
  31. Video in email
  32. A flash sale; how long should your redeem time be
  33. Benefit versus product feature driven
  34. Adding a Teaser email
  35. Personal tone versus business tone
  36. Sending a reminder
  37. Swap the order and position of your content categories
  38. Adding / removing editorial content
  39. Sending a triggered welcome email series
  40. The number of text links: a lot of links versus not so many links
  41. Top lists, or even list of lists
  42. Most popular items
  43. Using trust icons
  44. Removing links from your header navigation bar
  45. Using line of sight to direct the eye
  46. Different background colours
  47. Use humour or whit (Danger! Can backfire!)
  48. Newest / new in stock items
  49. Segment customer versus non-customer
  50. Test different navigation structures and designs
  51. Changing colours to highlight an important element
  52. Show spokespersons or ambassadors
  53. Using testimonials
  54. Use an interesting looking graph or flow diagram
  55. Make it look less like an offer / advertisement
  56. Clear versus teasing subject line
  57. Link to archives or related content
  58. Long copy versus short copy
  59. Use of bullet points
  60. Adding a footer navigation
  61. Testing the from name
  62. Use steps or a progress indicator in a series
  63. Add a variety of social proof
  64. Showing personal data (name, customer number)
  65. Loyalty points and customer levels
  66. Tie in to special days / events
  67. Being less lazy in your Call to Actions than “click here, read more”
  68. Repackage products into combinations and packages
  69. Removing clutter
  70. Repetition of the CTA and button
  71. Reuse last year’s successful campaign
  72. Change the writing perspective of the copy: He, Me, She, We
  73. Intellect Versus Emotion
  74. Write down your assumptions, then assume nobody knows
  75. Present a search box in the email
  76. A Call / Chat / contact now option
  77. Don’t sell to the ones that can’t buy
  78. Follow up on any downloaded content
  79. Change you Landing page design
  80. Present decision-required information (eg for an event, show date and time)
  81. Focus on the Greedy nature of the subscriber
  82. Change your incentives
  83. Rename (even if only in the email) your product / content
  84. Using a different designer
  85. Adding / removing index links
  86. Changing image Alt-texts
  87. Use the from name to show the type of email message
  88. Shopping cart abandonment emails
  89. Hook on to a popular trending topic
  90. Send more of previously successful campaigns
  91. Adding scroll indicators or scroll promoting design
  92. Ask to fill a wishlist
  93. Promote updating preferences
  94. Placement of Social Media buttons
  95. Highly personalised offers and content
  96. Call to action (CTA) button colour en design
  97. Add a see / search all catalogue link
  98. Customer versus non customer segmentation
  99. Segmentation based on engagement
  100. Introduce your team to the reader
  101. Make an unexpected offer
  102. Day to send
  103. Individualized Send Time vs. Universal Email Send Time
  104. Abandoned cart offer and timing
  105. Add a sense of urgency: “last chance, last dance”
  106. Equal or increased size for lead articles
  107. Reminder versus service update
  108. Behavioural data interests segmentation
  109. Social buttons design
  110. Think before you send
  111. Adjust your triggered emails by Season
  112. Personalizing images based on customer profile
  113. Splitting your CTA up into multiple, deeper linking CTAs
  114. Adding a PS
  115. The number of products in your mail
  116. Mail based on RFM scores
  117. Document the impact of the test on the funnel
  118. Coupon code versus a direct link mechanism
  119. Send Frequency and cadence
  120. Offering third party products or content
  121. Tone-of-voice: Human versus corporate
  122. Resending to non-openers
  123. A non-selling email
  124. Add / remove a contact center employee image
  125. Adding reviews or scores
  126. Product images versus people using the products
  127. Images of a successful outcome of using the product
  128. Different ways of segmenting your subscribers
  129. A mystery email
  130. Test call-outs, text pointing to particular parts of a picture
  131. Test violators, attention-grabbing shapes such as starbursts, ovals and banners
  132. Image heavy versus text heavy
  133. Market segment
  134. Inserting personal data in copy (name of business)
  135. Pintrest style email
  136. Use of first name in copy or subjectline
  137. Brainstorm more variants of a previous test
  138. Curated content versus original content
  139. Different CTAs inside product images
  140. Single message in your email versus multiple items
  141. Transactional email promotions
  142. Repeat your offer and main benefits on the landingpage
  143. Use of a Survey
  144. Using in-email banners
  145. Product versus product category
  146. Using different fonts and font sizes
  147. Mail based on engagement level (heavy opener versus never buyer)
  148. Doing Nothing
  149. Category landing page, versus product landing page
  150. Intro length
  151. Use of (previously) bought product
  152. Removing the intro
  153. Retest the test you did more than 3 months ago
  154. Pre-sales mails
  155. Email exclusive content
  156. Adding click indicators to your CTAs
  157. Loss aversion (don’t miss) in subject line and email
  158. Different type of offers (free shipping versus discount)
  159. Using a contest or prize draw
  160. Adding easter eggs
  161. Add an indication of reading time (eg approx 2 mins reading time) for articles
  162. Break one email into several more focussed emails or vice versa
  163. Adopt for preview pane reading
  164. For once: don’t offer a discount
  165. Add or remove a highlighted / featured article
  166. Move the email header and navigation down below your primary text/imagery and call to action.

A/B split testing your email newsletter

A/B split testing is a great tactic to optimize your messages. You know, there isn’t one right way to create, design and send your email marketing campaigns that will work for each brand and audience Not all of the above are suitable for your own email marketing program. But with that many variables, there are equally enough combinations of A/B split tests you can do.

If you are looking to improve your email marketing results, A/B split testing your emails is a great tactic and it should not be missing from your email marketing strategy. As you can see there is so much more to explore!

Let me know about your own split testing experiences, tips and questions in the comments.

About the Author: Jordie van Rijn

Jordie van Rijn is an independent email marketing consultant. Next to helping companies improve their email marketing results he gives email marketing training and is a writer and speaker in the field of online marketing.
  • Pete Austin

    Jordie asked me to comment on this. I’ll just say:
    (1) There’s not enough time in the world to do all this testing.
    (2) If it’s true that all of these 163 factors matter then clearly the influence of each individual factor must be too small to worry about.

    • Thanks Pete.

      I agree with your #1, there isn’t enough time. What I also see is that people start off enthusiastically and then run out of ideas and/or aren’t aware of the different options. I want to go on record and say that you should cherry pick the ideas from here. Definitely not try to do them all :)

      I don’t know the EmailReaction optimizer. Was it automated testing, how did that work?

      • Pete Austin

        The EmailReaction Campaign Optimizer was multivariate testing that was simple enough for a high proportion of marketers to use.

        They entered different versions of subject, content, etc. on a form and the system did all the work:

        It did sample sends using all combinations, then as soon as winners were identified it sent the overall winning combination to everyone else on the list.

        Most important – and uniquely in my experience – it got the math near enough right to achieve results that were significant enough to be useful from a small list size.

        • Nice, I was involved in building an external multivariate testing tool (via dynamic imaging) called emailtestbox. Although having it inside the ESP of course has more benefits.

  • Rut Dawson

    I’ve found adding / removing editorial content is one of the easiest tests you can do that can affect clicks in a big way. For a standard editorial segment on a digest / newsletter style email (headline / snippet / c2a text) – it can be lengthy if you’re including a number of posts.
    The snippet gives away a key part of the copy / article that you want the recipient to click through to read more about anyway, so a good A/B test would be this segment against a simpler headline / c2a segment and see if it makes a difference. If it’s successful, you can then lead onto increasing the number of headline / c2a links till you find an optimised number (or run out of space!).
    Some good ideas on this list to keep busy. Great stuff, keep up the good work Jordie.

    • That is a great test. So you would test if the text between headline and call to action for editorial content actually hurt or helped clickthroughs.

  • @Pete, I am little bit agree on your thought because if we go for all those test one by one , it will consume lot of time. And if we are working in large scale IT company, then it is not possible to do so. @Jordie, I am agree on your post also , because if we passed all this test, then Email campaign will rock. But i want to ask you something is that, is there any tool available which can test all these cases?

  • Kota Venkatesham

    Thanks for sharing such nice information.

    I have question is iam planning start new website for content mrketing .

    i am alredy hve 25+ selected products to promte through 1million bulk emils

    data base it work for me in long run please share your opinion

  • teaser emails are one of my favs! Thanks for listing these all out! Totally bookmarking this article! http://www.holisticfashionista.com

  • Wouter Krusemann

    would u agree that A/B testing is crucial? Sendinblue does not offer that option but has some other very promising option? What email client do you prefer?

    • Depending on your list-size Wouter, if it is big enough to get reliable results: yes testing is a surefire way to increase results (so you should do it).

  • Pingback: 150 A/B email split testing ideas - CommuniGator()

  • This blog post is really great; the quality information of this post is genuinely incredible.

  • Emma Jones

    Thank you for research on a/b testing. Really useful information. My service – SendPulse ( http://www.sendpulse.com ) – for a long time lured me on this function, but when I started to work in the context of email marketing, I underestimated the importance of a/b testing. It seemed to me that I can predict the demand of a varinat letters by myself. Now I also use a/b testing before each mailing and very happy with the result.

  • Pritha Bose

    Thanks for the research. In most of my email campaigns, testing the call-to-action button and placement of unsubscribe button works best for me. Although I must say sometimes even tweaking these two does no good, and I need to create a whole new template from scratch. Your research definitely has given me ideas on what else i can try to test and create a better engagement. So, thanks!

    • Yes, most people will start with optimizing simple elements like subject line or CTA, but there is so much more that can have massive impact.

  • Pingback: Newsletter Design: 10 Awesome Hacks For Emails That Convert()

Hi there,
Want to get more out of email marketing? Subscribe, because the really good email marketing tips are in the newsletter.
You can unsubscribe anytime, but you won't even want to.