Lost in the Sea of Customer Experience? Harness the Power of Experimentation

Imagine falling overboard in the middle of the night into the vast Indian Ocean due to acute seasickness and then spending the next 29 hours wondering if anyone would ever find you? Even with a few things on your side - warm tropical water, for one, and the fact that you’re a strong swimmer, familiar with the ocean, for another. It doesn’t matter much as you watch the boat you just fell out of disappear into the night. Knowing what needs to happen, doesn’t make much of a difference when you have zero control over the circumstances.

This was the true life experience of Brett Archibald, a Cape Town (South Africa) businessman who was lost at sea for 29 hours while on a surfing vacation in Indonesia in 2013. It was past midnight when he went outside and leaned over the edge of the boat to get sick. He didn’t even know he was seasick enough to pass out, until he hit the water with a splash. He spent the next 29 hours trying to stay afloat, while warding off sharks, seagulls, and overwhelming feelings of despair. These were made worse when on two occasions he saw boats coming close and then veering away in the opposite direction. 

But there’s an interesting twist to Brett’s story. While he was drifting, a massive search was being launched by local rescue services as well as several privately owned boats. The area where Brett had fallen overboard was en route to a small group of islands known as the Mentawai. In this area, the current typically ran southwards and the boats involved in the search were using that assumption to map out search patterns. But an Australian surfer, Tony Eltherington, who had joined in the search with his private boat, noticed that this wasn’t in fact the case. By throwing coconuts overboard at intervals and watching them drift, he determined the currents were actually flowing in the opposite direction - northwards. This meant that the currents were taking Brett in the opposite direction from where the search was being conducted. With this new information, a new search area was mapped out. Two hours later Brett was found, weak, sunburnt, and dehydrated, but alive.

What does this have to do with optimizing CX?

This story highlights how assumptions can steer teams in the wrong direction and hinder efforts to find the experiences that customers want and need. Too much reliance on historical customer data or trends can create bias, meaning ideas that could make an impact on customer experience (CX) are discarded, because they’re assumed to not align with past trends or knowledge. 

That’s not to say that there isn’t any value in historic customer data or trends. But it needs to be viewed in context. With customer preferences changing so rapidly, learning from real-time interactions is a far better indicator of what experiences resonate with visitors. Especially when understanding how and why consumers shop is constantly evolving. 

Being able to learn from how visitors are browsing a website and what experiences are progressing them further along the customer journey, also highlights when shifts in behavior occur. Having the ability to identify shifts quickly, helps brands respond faster and be able to continue to offer engaging and relevant digital experiences.

How AI-driven experimentation improves CX

In the story of Brett Archibald, he was found because Tony Eltherington decided not to rely on historical data but rather he experimented to see which direction the currents were actually  flowing. Some might have viewed his efforts as foolhardy, after all, everyone knew the currents in that part of the ocean typically always flowed from north to south. There was a possibility the experiment may have rendered a result that was already known. Except it didn’t. On that day they were flowing northwards. Had Tony not run the experiment, that information would have remained unknown.

Applied to optimizing CX, where buying behavior is constantly evolving, it highlights the need for ongoing experimentation. When this approach to optimization is paired with technologies such as artificial intelligence (AI) it becomes a powerful weapon for brands to increase conversions, improve CX and gain a competitive advantage. 

AI-driven experimentation is able to follow what’s happening on a website and learn from real users in real-time. It hones in to identify which ideas visitors are responding to and adapts the experiences being served next, based on these learnings. 

A huge advantage of AI-driven experience optimization is the ability to adapt or add to top performing ideas and improve digital experiences even more while the experiments continue to optimize for the best CX. Additionally, low performing ideas can be paused so that the experiments focus on identifying the ideas that are getting the best response from visitors. It is a distinct departure from traditional A/B testing tools where variables can’t be changed mid-experiment. Instead, AI-driven experience optimization enables continuous experimentation and personalization to keep pace with changing consumer behavior and trends.

AI-driven experience optimization uses experimentation to serve up hundreds of different experiences to website visitors working from different variables in thousands of different combinations. While this may sound hugely complex, for AI this task is effortless. AI thrives on complexity, the more  data available, the more accurate the insights generated, and this is what makes it so valuable. Accurate insights, generated quickly, enables brands to quickly pivot and respond to changing customer preferences.  

The reality is that the task of optimization and improving CX is never done. Even once insights are obtained, it indicates what has happened and what customers are likely to respond to, but  doesn’t guarantee that  they will in fact respond in the way expected. The learnings from today can be used as a baseline to improve CX tomorrow but it’s still a constantly moving target. This is why being able to deliver a consistently good customer experience requires ongoing experimentation. Each new learning builds on existing knowledge with the aim of keeping pace with changing consumer buying habits and expectations.