In this special guest feature, Shana Pilewski, Senior Content Marketer at Dynamic Yield, takes a look at retail coupled with the benefits and challenges with adopting today’s leading technologies like AI and machine learning. Specifically, the key to humanizing experiences and influencing action is to treat each outcome as unique and dynamically respond to each customer individually, a feat which can only be scaled with machine learning. Shana is Head of Content at Dynamic Yield where she leverages her skills in content strategy, development, and creation to help marketers level up their personalization skills.
Even in a time of declining early-stage VC investment, nearly $5B was pumped into artificial intelligence startups in 2016. The most abused buzzword of 2017, everyone seemed to be talking about “artificial intelligence” and “AI,” but a lot was left to the imagination when it came to actual adoption. Despite the hullabaloo around AI, augmented reality and other futuristic technologies, little old personalization ended up the top inquiry made by retailers into leading analyst firm Gartner.
However, the problem with adoption of AI in industries like eCommerce has not been due to limitations of the technology itself. In fact, while eCommerce may not commonly be thought of as a hotbed for AI innovation, there are scores of machine learning startups that serve the sector. eCommerce giants Amazon and Alibaba are already using AI to implement end-to-end solutions focusing on the overall retail experience. And even for the rest of us, low hanging applications of artificial intelligence, albeit less sexy, can still support a desire for providing consumers with a more human touch. For example, identifying which recommendations strategy to serve new, returning, or loyal visitors to an eCommerce website.
But in order to meaningfully impact the customer experience on this one-to one level, machine learning algorithms need to ingest vast amounts of data. Much like an in-store associate collects information from a customer about what they are looking for, brands need to piece together clues users leave behind to better serve them. And the more data sources, the more successful an algorithm will be at predicting the desired outcome for each user. Trouble today is that even for tech-forward retailers, this data often exists across a myriad of siloed software programmes, making it impossible for them to create unified profiles of their customers for powering truly individualized recommendations, and experiences beyond.
Only with an understanding of all the actions a user took before interacting with a brand, can the modern customer journey be tailored to the person, and we all know that requires more than a sliver of data. It means understanding which homepage banner a user clicked on, how they arrived on a product page, and what incentives they commonly respond to. Then, using algorithms that constantly collect all user data and signals, serving the best variation to each individual user in real time, regardless of where they arrive from, what device they are using, and so on. AI not only rids eCommerce marketers of infinite tedious data analysis, it maximizes revenue generated as it optimizes more quickly than the average A/B test.
The key to humanizing experiences and influencing action is to treat each outcome as unique and dynamically respond to each customer individually, a feat which can only be scaled with machine learning. So while AI won’t upend retail anytime soon, practical applications of machine learning to common eCommerce problems will proliferate in 2018, benefiting brands that adopt the technology and a more personalized approach.
Sign up for the free insideBIGDATA newsletter.