You probably spend months preparing and strategizing for the holiday season, since for many retailers, it's the most lucrative several months of the year. Split testing is one of the most accurate ways to gain insights for what strategies you should employ each year. We know that eCommerce is always evolving; what worked last year might not work this year. You can use A/B testing to understand what works best for your business and your customers.
After reading some tips below, set up your own split tests right in the Listrak application. You have two options for setting up split tests:
ASK YOURSELF THESE QUESTIONS FIRST
Tests should be structured in a way that help drive business decisions and results. Always begin with a strong hypothesis and goal, which you can clearly define by asking yourself the following questions:
- Specifically, what are you trying to learn?
- What are your assumptions?
- What variants will you test?
- Are the changes big enough to truly make an impact?
- Will the tests lead to improved results?
- How will you measure the results?
- Can you re-use the results to inform future campaigns?
- Will the results answer your hypothesis?
BE MINDFUL OF TEST CONSIDERATIONS
- Build split testing into the campaign development process
- Set goals
- Make sure to stay on brand
- Don’t test random elements
- Make sure your sample group is large enough
- Measure winning campaigns by the right metric
- Don’t call tests too early
- Test what matters and what will make a difference
VISUAL SPLIT TESTS
- Product image variations
- Product image vs. people using product
- Product image sizes
- Headline size/header height
- Variation of color scheme
- One column vs. two vs. three
- Template (current vs. new)
- Use of price ribbons
- User Generated Content (UGC) imagery
- Use of trust icons
- Navigation structure/placement/style
- Social icon treatment/placement
- Personalized images based on customer profile
- Image heavy vs. text heavy
- Call to action (CTA) design/style/indicator/size/location
- Animation (animated gifs) and/or video link
- Design style
- Graphic order/scrolling (top vs bottom engagement)
CONTENT SPLIT TESTS
- Subject line
- Pre-header text
- Headline/body/CTA copy
- Personalization (use of customer name)
- Content based vs. product based
- Pricing and discount variations (10% off vs. $10 off)
- Negative pitch (loss aversion)
- Content/link flow/order
- Product description copy
- Word variations (e.g. free vs. complimentary)
TIMING TESTS AND MORE
- Time of day
- Time of week
- Pre-sale teaser/announcement
- Flash sale
- ‘From’ name/address
- Segmentation (based on engagement, customer vs. non-customer, etc.)