The A/B Testing Secret No One Warns You About

The retail world is decidedly digital. Our clients have shared a challenge it is to manage product data content at e-commerce retailers to ensure that it is compliant, effective, and helps drive sales. We hear you! And we have chronicled your challenges in our e-book that addresses Ten Pain Points in E-commerce to Overcome.

 Challenge #9 is A/B does not equal Why

A/B testing may get you the ‘what’ but not the ‘why.’ How do you diagnose what makes your product images successful?

As we reviewed in the last post, A/B testing, which some have considered as the optimal solution for measuring the effectiveness of digital assets, is not without its challenges when it comes to the digital shelf.   

 Even after you factor in all the ancillary elements that might impact the outcome of a multi-week or even months-long test, there is a massive blind spot: I might know that one SKU in the A/B test performed better than another?  But why?  What do I change?  What do I fix? 

 Price might be unambiguous, but even price examined in a vacuum doesn’t reflect any other benefits and barriers that are affecting shopper choice.  

 As we covered in other posts, e-marketers do have guidelines, however ambiguous, for much of the text-based content included in an average PDP, which differs by e-tailer and category. The most perplexing and unhelpful recommendations are the guidelines regarding images that admonish marketers to make pictures “compelling’.  

Our advice is straightforward from a behavioral perspective: the critical goals for any content on the PDP, especially images, is that they must communicate a definite benefit, making a consumer’s choice easy, and the photos must eliminate barriers and avoid causing friction that would stop a shopper from selecting your product. And those behavioral elements can be measured and rendered as scores that are predictive of a product ending up in the digital cart. 

 Decades of learning have helped us create behavioral metrics and an extensive database of pack and shelf designs in physical retail. We measured attributes like viewability, findability, and shopability (Would the item actually be purchased or selected?). Many of the same behavioral principles apply on the digital shelf relative to the prediction of purchase.

Therefore, we can unambiguously measure the potential success of PDP images (that are the proxy for a shopper experiencing the physical product in a brick-and-mortar retail environment) with confidence that they will drive choice in e-commerce sites. How we measure, compare, and score the PDP images on the digital shelf, using image recognition AI, adds the “whys” that A/B testing misses. 

 What an e-marketer might want to know about images in order to optimize them should fit in the measurable categories like  

  • Does the image convey the perception of quality?  
  • Does the image communicate product benefits? 
  • Does it influence behavioral buying? 
  • Does the image provide additional helpful information to facilitate selection? 
  • In specific categories, does the image communicate that the product would taste, smell or feel great? 

With a large enough database of previously tested e-commerce images and a robust image recognition algorithm, the visuals in question can be compared and contrasted to norms; a predictive score can be derived against these (and other) characteristics to identify and prioritize images that really need attention because they are creating a barrier to shopper choice.   

Here is where the real improvement over A/B testing comes in: a comprehensive database could also suggest edits for deficient images. These suggestions can be as explicit and straightforward as “Emphasize branding and font size to bolster quality and benefits communications” or “Emphasize people interacting with the product to establish size and scale.” These explicit instructions provide the busy e-marketer with a roadmap for optimization and know precisely the “why’s” and what to do!  

Finally, a powerful analytic engine parsing data at the speed of AI can provide an always-on monitoring system delivered in an easytonavigate dashboard interface, with highly visual renditions of the scores, providing these diagnostics virtually in real-time rather than the weeks or month an e-commerce A/B test requires.  

But is this theoretical? Or is there really an affordable technology-driven solution that comes from a trusted company with genuine domain expertise in shopper marketing that brands can apply to the challenge of images in e-commerce environments today?  

The Solution?  Flash.PDP™!

Behaviorally is the leading digital partner to help brands drive shopper growth. Knowing these challenges exist for our clients on e-commerce teams who want to win in digital retail, we developed a solution that leverages visual recognition AI. Our extensive database of shopper marketing content, our unique behavioral framework, and decades of category expertise.  Introducing Flash.PDP – an always-on alert system to identify and optimize product images on the PDP that will convert to sales and drive shopper growth.  It addresses category and retailer-specific metrics that provide easy, efficient ways to monitor and optimize images to increase sales.

To learn more, contact a Behaviorally digital retail expert today here.

We are Behaviorally, The Transaction Experts

We Define and Diagnose consumer behaviors that create valuable transactions.

Contact Us
photo_2x2