In the future, AI will be everywhere

Science fiction writers have seen it coming for decades, naturally, but the real world is definitely catching up: AI is now a $15 billion industry, and is expected to be a $70 billion one by 2020.

Since AI gets everywhere, it’s no surprise that it’s infiltrating the user experience (UX) field, which in retail helps consumers streamline their shopping experience. With automated learning, a site can know which product to put in front of which customer. In short, AI can personalize a buying experience, and retailers are increasingly realizing this.

Ecommerce sites optimized for UX have seen 70 percent more products sold alongside a 90 percent reduction in support costs. I chatted with Barry Pellas, PointSource CTO and Chief Business Technologist, to learn more about how it works and why context is essential. Here’s our Q&A.

So how do you use AI to improve UX?

Technology’s role is to enable the user experience and ease the barrier of entry for developing deep personal experiences with applications. When AI is incorporated into the picture, UX teams are provided with the ability to utilize this new toolset and create more engaging experiences. AI can learn user behavior and enable us to engage users on a personal level based on how they prefer or plan to utilize an application or system. AI can also predict what will happen based on the users’ context so the next action is aligned with the user’s intent.

Do you have any examples?

This could be something as simple as a UI that takes input generated by an intelligent system to show suggested next actions or quick actions based on your personal history of those actions. You can see this on platforms where quick replies are based on how you typically reply to messages or mail. In that case, instead of suggesting “OK” all the time, the system knows that you are more likely to reply with “K.”

Another example might be when creating content for users to consume. AI can help content creators by allowing them to use voice search or natural language to find images without having to do traditional searches. Authors can simply ask the system, “Can you give me some examples of images that contain A in the context of B?” and the system will reply with images that match that query from the natural conversational language.

Does this work best for certain types of products than for others?

It doesn’t have to if the context is known to the AI. For example, let’s take making an online purchase for a football jersey that has been customized with a name, number, etc. If the AI knows that where the user will play games during the Fall tends to be cold and rainy, then the system can recommend gear based on the user’s context.

In addition, AI can utilize the data from conversion analytics to break down barriers and fast track purchases, increasing conversion rates. In this case, if the user has bought around this season before, the AI will know the actions they took and the times they have interacted with the retailer. Since it can identify the context in which the user interacted with us previously, it would change to a quick flow method for the user to purchase the gear they need for the season with very few clicks.