AI in Commerce: Takeaways from The Edge Summit
In a soon-to-be released survey conducted by Blue Yonder, 97% of the retail executive respondents stated it was a priority to adopt AI for omni-channel applications. That’s a big number, but it’s also not surprising.
I just had the privilege of attending a timely and unique event on how to “Shape the Future of AI-Powered Commerce.” The Edge Summit event was hosted by Bloomreach and brought together a stellar lineup, including speakers from Wayfair, StitchFix, Canadian Tire, ThirdLove, and Total Wine, all to talk about the impact, priorities and the future of artificial intelligence (AI) and generative AI – specifically around commerce. It also aimed to look at insights into where retailers are already getting started with AI and where they see the near and far opportunities.
My four biggest takeaways from the event are:
1. Generative AI is a Game Changer and Priority for All Retailers
As reflected in the soon-to-be released research from Blue Yonder , AI and generative AI is a priority to understand, learn, and apply to retail use cases. The speakers and the attendees all likened this moment in AI adoption to other key moments in time – it’s 1997 and the internet is gaining traction, its 2006 and the smartphone is revealing its power, and, now, it’s 2023 and the curtain is going up on Act 1 of the generative AI play. This technology is a game changer, it’s powerful, and it’s exciting and scary – but at the same time it needs rules, governance and frameworks, as well as the human element.
2. It’s a Together Thing: Humans and AI
One of the major themes repeated over the course of the conference was the need for humans and AI to work together. First, AI should augment human decision-making, including assisting employees with making better, data-driven decisions and improving the customer experience by using insight into who the customer is, what makes them tick, and what they need to provide recommendations or influence purchases. Outside of customer service, there is enormous potential for AI to play a role in optimizing the supply chain, creating efficiencies in inventory placement, and driving to more sustainable approaches to enhance fulfillment. Another key element to humans and AI working well together is visibility into why AI makes a specific recommendation or suggests a specific action. At Blue Yonder, we talk about looking into the ‘Glass Box’ – not the Black Box.
3. Conversational AI: Customer Service are Top, First Applications
By and large, almost every retailer to discuss initial uses of AI in their business referenced using generative AI in their customer service. First, personalization and understanding the customer becomes exponentially easier and more insightful with generative AI. Every move a customer makes across their shopping and searching journey can be fed into large language models (LLMs) and AI models to gain better understanding of the individual consumer and the consumer as part of various cohorts. Then, use this data to enhance the recommendations and actions retailers take to better serve the consumer. Use cases range from serving up better product recommendations to enhanced and augmented customer service. Conversational AI allows consumers to engage in natural conversations with chatbots and customer service modules for self-service, order modification, product inquiry, and more.
4. Beware of Hallucinations
Perhaps I was in the dark before, but a key learning I had from the event was around the danger of AI hallucinations. What is an AI hallucination? No, they aren’t drug-induced, but they do create a false reality. Basically, the data feeding the open AI systems (e.g., Chat GPT) are from public internet sources like Reddit and Twitter and that means the data accuracy is based on the goodwill and accuracy of people and their opinions. In every aspect of business, we have always learned that the information and insights gleaned from the data can only really be as good as the source and accuracy of the data. When it comes to designing and creating generative AI capabilities for business, creating a private and safe LLM – where the data feed is reliable – can make a big difference in the results and the impact of the solution on your business.
All in all, The Edge Summit was a wonderful event, in a beautiful location, with great attendees and networking, and insightful content that hit at exactly the right time. Thank you to Bloomreach, the speakers, and everyone I met at the event – I look forward to working with you soon!
As another preview to the soon-to-be released Blue Yonder survey, almost all retailers (95%) are currently utilizing AI in at least one supply chain or order management-related application or use case. Among these retailers, generative AI tools (like ChatGPT) are the AI-powered technology being used most. Despite being at the beginning of Act 1 of AI theater, many retail executives feel good about their understanding of generative AI and perhaps feel less good about predictive AI. Stay tuned for the results and findings of our Blue Yonder survey of 150 retail executives and their take on AI in commerce.