It’s that time of year when experts weigh in on what the next year will bring for retailers. Rather than listing the top retail technology trends like I did last yearI’m going to focus on generative AI and highlight some key trends to watch. If 2023 was the year that generative AI exploded onto the scene and 2024 was the year of experimentation, then 2025 will likely be the year that the technology matures even further. We’re getting closer to Gartner’s “Slope of Enlightenment,” but we’re not quite there yet.
If you’re tired of hearing about generative AI, I can sympathize. Generative AI reminds me of the band Nirvana. Many grunge bands blazed the trail, but Nirvana got all the attention at the expense of many other big bands. But no one can deny that Nirvana and generative AI have had a major impact on our world.
Here are three use cases specific to retail and three technologies to watch in 2025:
- Virtual shopping assistants (use cases)
- Hyper-personalization (use cases)
- Virtual try-on (use case)
- AI agents (technology)
- Domain (technology) specific foundation models
- Computer use (technology)
Virtual shopping assistants
The concept is simple. When a shopper doesn’t know what to buy, they can ask store associates for advice, at least in theory. But what happens if you shop online? Enter the AI-powered virtual shopping assistant, mastering topics as diverse as plumbing, digital networking, and cold-weather fashion. What tools do I need to repair an in-ground sprinkler system? Which Wi-Fi router works best outdoors? And what should I look for in stylish ski gloves? Getting answers to these questions was the idea behind Rufusthe virtual shopping assistant launched by Amazon.
What makes Rufus particularly useful is the fact that it is a conversational tool, allowing a back-and-forth with buyers until they are satisfied with the responses. Just like a human expert, the virtual shopping assistant asks questions to help understand the buyer’s needs and preferences. Some might consider this conversational research.
It certainly doesn’t replace traditional search, so retailers should start by modernizing their product search and discovery solution and then decide if chatbots and virtual assistants make sense for them. These solutions will likely increase buyer confidence, leading to increased sales and possibly fewer returns.
Hyper-personalization
Personalization using machine learning has been used for 25 years, when Amazon started using collaborative filtering which predicts the preferences of a given customer based on the likes and dislikes of similar customers. The next wave combines machine learning and generative AI to create individualized experiences for shoppers. This includes hyper-personalized marketing communications, search results, product detail pages and even chatbot conversations.
Ultimately, each online store session could be personalized for individual shoppers, presenting them with products using engaging themes, tailored assortments and curated offers based on their personal preferences. This type of white glove service, once reserved for the wealthy, could trickle down to average shoppers. Retailers need to think about how best to hyper-personalize every interaction with shoppers, leveraging data like past sales, product data, and third-party customer data.
Virtual try-on
Online sales have been hampered by a lack of buyer confidence, particularly in areas like fashion. Without being able to conceptualize the product, they may be reluctant to purchase it or order multiple versions and return the unselected ones. Generative AI opens up the possibility of visually representing products in context so shoppers can virtual tryout products. This is achieved by combining two images, for example a person and a sweater, or a chair and a living room, so the buyer can evaluate the look.
AI models like Stable Diffusion and Amazon Titan Image Builder are used to intelligently combine images, showing buyers what to expect and increasing their confidence when purchasing. Retailers selling clothing, fashion, accessories, furniture, or other products that benefit from visualizations should consider this feature.
AI Agents
Chatting with a chatbot or virtual assistant can be informative, but is rarely action-oriented. Agents, on the other hand, play a role in achieving a goal. They are generally self-sufficient and have tools that help them accomplish specific tasks. You can even view agents as part of your team, contributing and getting things done. Products like Amazon Bedrock Agents can use chain-of-thought reasoning to break down and solve complex problems. For example, you might have a pricing agent who can research prices on competitor websites, review product margins, and make price recommendations within its defined rules.
Imagine that when you buy forecasting software, for example, it comes with a forecasting agent who can operate the software on your behalf, updating and distributing your forecasts as needed. Retailers should look for tasks that could be automated with agents, allowing for greater overall team productivity.
Domain-specific foundation models
Most basic models (FM), like large language models (LLM), are trained on a corpus of public data in order to have general knowledge, but it is also possible to create a model from scratch to learn more. focus on a particular area. The Amazon Science team created a Retail specific LLM for use by Rufus which has been trained on its extensive product catalog, customer reviews and other similar data with the aim of improving the shopping experience. The hope is that its focus will allow it to be smaller and therefore cheaper to operate while still producing superior results.
Of course, creating an LLM is a massive project and probably out of reach for most retailers. Most will therefore choose to refine existing models with their own data. Retailers should consider this cost-effective approach to improve generative AI results.
Computer use
Although it is still in its early stages, it is possible to leave an FM, like Claudius 3.5take control of your computer and use it the same way you would use it. For example, you can ask it to create a purchase order for you, then it will “look” at your screen, take control of your mouse, and fill out the form. Generative AI can also be used for regression testing, ensuring that changes to your online store work as intended. On the consumer side, a buyer could request to find and purchase the cheapest Apple AirPod Pros, for example, letting them do the research and ultimately select a product to purchase.
As FMs become trained to use software, particularly browsers, they will be able to automate some routine tasks, allowing humans to spend more time on creative activities. It’s a little too early to adopt this technology today, but keep an eye out for general availability in 2025.
Visit AWS and NRF
To help inform your decisions about adopting generative AI in 2025, we invite you to visit the AWS booth at NRF in January 2025 to speak with experts from AWS, Amazon, and our partners, explore exciting demonstrations of generative AI and discover the latest ideas. in this area, in addition to many other innovations in the areas of smart store technology, digital commerce, retail operations, etc.
Additionally, here are some great resources for tracking the progress of generative AI in the retail industry:
