By Bill Cunningham
May 1, 2018
This week’s selection of articles span some terrain, but have important attributes in common. Each article focuses on how we can use A/B testing and research to understand customer behavior in a way that data on its own cannot. Read on to learn about ways to predict A/B tests in order to even further maximize your CRO program, how AI does not necessarily lead to improved customer service, and how research tells us how readers interact with short articles (meaning you can then run A/B tests with more directed goals).
Author Jakup Linowski presents his research here on predicting winning A/B tests. By finding patterns in winning A/B tests, Linowski was able to demonstrate a 71% success rate in predicting A/B test results. These results are substantially higher (up to 80% success) in relation to higher degrees of repeatability. Identifying patterns to accurately predict results adds a layer of control and efficiency to A/B testing, potentially leading to even better optimization. Rather than A/B testing features on your organization’s pages with no sense of the result, this research suggests that you could specifically A/B test features that you were pretty sure would improve your KPIs. That means more efficient A/B testing, more conversions, and better customer experiences all with the same amount of A/B testing.
The Takeaway: There are predictable patterns from A/B tests that can help you systematize your A/B testing process and maximize conversions and cx.
Why AI Can’t Fix Your Broken Customer Service Model, VentureBeat
In this article author Dr. Vinod Vasudevan dives into a customer service case study. After a flight gets cancelled due to bad weather, a customer gets terrible service in regard to rescheduling a flight. He asks the question: would AI mean better customer service in this circumstance? Vasudevan’s answer, shortly, is no. Good customer experience stems from a brand’s attitude and philosophy toward customers. AI can help with that — it can offer a kind of small business personalization to a huge user base, but it can’t prioritize service. He writes, “While artificial intelligence can deliver better customer engagement and, ultimately, better profits, it cannot make up for a company not being customer-focused. That’s a human problem, not a technology one.”
The Takeaway: Good customer service stems from a brand’s attitude toward customer service, and AI can’t fix a bad cx attitude.
In this article Madeleine Sidoff presents her original research on how customers read short articles online. Since content marketing has become the name of the game, this information is crucial in understanding how your customers interact with your brand. Sidoff’s study split users into two groups, one aged 18-30, the other 50-60. In this study, it seemed that all users read the headline, the first half of the article, and the featured image. Only 62.9% of younger readers read the entire article and 54.5% of older readers read the entire article. Although, 91% of all readers read the image caption.
The Takeaway: If you’ve made it to this part of this blog then you are in the minority of readers.
A Little Point of View
While the collection of articles this week is somewhat diverse, they each offer important insights into really understanding your customer to make the most of optimization solutions. While A/B testing will always give you valuable information, really learning about your customer can help you design more intelligent A/B tests for an even bigger boost to the bottom line. The trend this week is all about marketers not only collecting and utilizing data but using numbers to understand user behavior in a more human-centric way. Artificial intelligence, automated A/B testing, and machine learning are all incredible tools. But, they don’t replace human interpretation. To get the most of your A/B testing solution and conversion rate optimization, use all the resources at your disposal to then interpret your customer, and put their experience at the forefront of your brand strategy.
To learn more about SiteSpect, visit our website.
Subscribe to our blog: