By Leon Gauhman, co-founder and CPO/CSO at digital product consultancy Elsewhen
From AR to embedded finance to delivery drones, retail bosses are constantly assailed with promises around radical tech disruption. But Generative AI, specifically Large Language Models (LLMs), represents the industry’s most seismic opportunity yet.
OpenAI’s ChatGPT Enterprise and Google’s soon-to-launch Gemini AI are potential game-changers for retail brands, providing the elusive link between insight and action. For a vertical that relies on its ability to interpret and unlock data, one of the most profound benefits of LLMs is their scope to deliver unstructured data analysis that can decode consumer behaviour at an individual level.
In retail terms, leaning into LLM’s capabilities means planning, launching and adjusting hyper-personalised experiences rapidly and at scale. Retailers who seize the challenge can expect to elevate brand experiences to a whole new level of impact – unleashing a new chapter of mass personalisation that would be hard to imagine, even a year ago.
So, what does this AI-powered vision mean in practice? Here are five ways LLMs can enable retail’s tech pioneers to reach new heights in consumer experience:
The rise of customisation experiments using unstructured data: Retailers hold valuable data across multiple systems. Until now, they have needed to spend heavily in cleaning and structuring this data, often via narrow-scope customer data platforms, before extracting value from it. Large Language Models available through Gemini AI and ChatGPT Enterprise can effortlessly interpret unstructured data from different sources. This creates a pathway for retailers and brands to run customisation experiments based on user likes and preferences. These experiments can help create bespoke experiences for individual customers at scale based on an unparalleled, data-driven understanding of customers.
Hyper-personalised rewards to boost engagement: Using LLM-powered data analysis, brands can develop hyper-personalised reward programmes peppered with customised surprises and rewards. Tesco is already deploying Generative AI to shape personalised customer rewards and promotions based on their behaviours. This benefit extends to how retailers communicate with customers, using Generative AI/LLMs to create copy assets code in highly tailored creative campaigns.
Conversational shopping at scale: With ChatGPT now able to “see, hear” and conduct conversations, retailers have a new route to developing even closer, more human customer relationships. US supermarket giant Walmart is already leveraging conversational AI to elevate its Walmart Voice Order feature, activated via customers’ smart speakers and mobile devices. The system uses LLMs to understand voice requests (e.g. “add eggs to my shopping cart”), pairing them with previous shopping information to predict brand preferences.
LLMs can also help with translation. Walmart has used AI tools to improve its chatbots, training them in localised phrases in markets including Chile, Mexico and India. This means it can better understand users’ real-time conversations and needs leading to improved customer satisfaction scores.
Personal shoppers for everyone: Through bespoke services such as Stitch Fix, personal shoppers are on the rise. LLM’s capabilities mean the end to shoppers’ time-consuming searches through a retailer’s inventory to locate the products they want. Instead, shops will offer AI-driven tools or assistants that provide a deeply personalised, enjoyable and easy service previously confined to elite outlets like London’s Savile Row tailors. Customers can chat with a retailer’s inventory, enjoy highly customised recommendations and have deliveries and returns organised for them, hassle-free.
Developing synthetic personas to design personalised products: LLMs are key in creating synthetic personas that help test hyper-personalised products and services. This eliminates the need for lengthy, expensive focus group testing while speeding up new product timelines.
While synthetic personas are at an early stage, brands such as Sephora are already using Generative AI to analyse deep-rooted customer data, which can generate briefs for new products tailored to specific customer needs. Likewise, Levi Strauss and Co. envisages a future combining digital design, mobile and online selling insights to manufacture products to meet precisely emerging consumer demands.
Final thought
The days of mass-produced retail experiences that rely on consumers to do the hard work, such as combing through a brand’s inventory, are at an end. The power of LLMs and Generative AI will drive a new era of mass personalisation that will change the face of commerce forever. Instead of being a chore, shopping will be a hyper-personalised, natural experience closely aligned with consumers’ needs. While welcoming this AI-driven retail renaissance, it’s also crucial that we guard against the negative consequences of unregulated and intrusive tech innovation. Data protection legislation and the use of trained human supervisors will be vital in developing an LLM-based retail ecosystem we can be proud of.
