Is GenAI the “Next Big Thing” in Retail?
The retail industry is moving fast. Evolving customer expectations and needs, fierce competition, and the quest for enhanced online customer experience – among others – are all factors driving retailers to rush into experimenting with emerging technologies. In 2022, newspapers were crowded with titles of bold retailers and brands landing in the Metaverse; 2023 has already proved to be the year of Generative AI (GenAI).
Following the hype, in the last months we have seen the web crawling with OpenAI’s DALL-E experiments, an example is the viral AI-generated video series “By Balenciaga”, or the ultra-realistic image of Pope Francis wearing a Moncler puffy jacket. However, the potential applications of the technology in the retail industry are much more tangible than digital content creation for mere customers’ entertainment.
GenAI technologies (such as ChatGPT and Dall-E), and solutions powered by LLMs, or text-to-image models are publicly set to have a disruptive impact across all retail operations and functions. The expected benefits of the technology span from enhancing and accelerating retailers marketing creative process to achieving the personalization of online customer journeys based on individual preferences and needs.
How Are Retailers Approaching Generative AI Nowadays?
With the OpenAI release of ChatGPT in early November 2022, large language models (LLMs) or foundation models (FMs) become available to the general public, and everyone had the chance to experience firsthand their transformative power in terms of augmenting human productivity and creativity- retailers included.
According to our findings, 40% of worldwide retailers and brands are in the experimentation phase of GenAI, trying to figure out the most relevant field of applications and use cases, while 21% are already investing in the implementation of GenAI technologies. Today we see organizations such as Coca-Cola, Mattel, and Carrefour starting to pilot GenAI applications, even if still on a limited scale and with a test-and-learn approach, predominantly across the areas of product development, marketing, and customer service.
Download eBook: Generative AI in EMEA: Opportunities, Risks, and Futures
Which GenAI Use Cases Do Retailers Expect to Explore in the Short Term?
In the short-term, most retailers expect to explore LLMs and FMs applications in marketing, sales and customer engagement, as confirmed by our data showing that 50% of them will prioritize GenAI marketing use cases over the next 18 months. Knowledge management, design and conversational applications are following marketing in the current retailers’ prioritization of GenAI use cases, while code applications are lagging for the moment- as opposed to other industries where they are leading the way for the technology implementation.
Unsurprisingly, retailers are starting to approach GenAI through less strategic and data-sensitive use cases using open-source commercial models or plug-in solutions (i.e., marketing and conversational applications). Other more strategic retail use cases (i.e., product prototyping and knowledge management) would require the fine-tuning of commercial models to the organization dataset and eventually the training of their own proprietary model or enterprise solution.
Therefore, we could expect these strategic use cases to reach wide adoption in the medium-long term- when the technology will become more established, and retailers will have developed enhanced GenAI in-house capabilities and workforce change management.
How Will GenAI Impact the Retail Industry?
We define Generative AI as a branch of computer science that involves unsupervised and semi-supervised algorithms that enable computers to create new content using previously created content, such as text, audio, video, images, and code in response to short prompts. Therefore, the power of GenAI goes beyond chatbots and lies in its ability to create infinite, contextualized content of any format.
Theoretically, the Retail industry could see significant improvements when advanced GenAI’s content creation ability will be added on top of the existing AI/ML applications that retailers use to automate processes, enhance personalization, and increase efficiency- and across all retail functions.
In the short-term, even if the technology will inevitably raise important questions in terms of proprietary data sharing, customer data privacy, and factual inaccuracy, without any doubt the integration of GenAI into online customer journeys could lead to significant improvements in both the backend of e-commerce workflows and the front-end of online customer experience– leading to major enhancements in retail e-commerce efficiency and efficacy.
The disruptive power of GenAI for digital commerce is not going unnoticed, for instance, the Chinese e-commerce giant JD.com announced the imminent release of its own retail-specific ChatGPT solution which aims to improve online retailers’ rankings of product listing on SERP, generate product descriptions that are tailored to a shopper’s preferences, and optimize online product images and video generation process.
Watch the Webcast: Generative AI in EMEA: Opportunities, Risks, and Futures
Two Nascent GenAI Use Cases with Examples
Conversational Customer Service Chatbots: Carrefour
Compared to the standard chatbots that are typically limited by a defined number of interactions and decision trees, GenAI conversational applications open an infinite number of responses to potential customers’ inquiries. In doing so, the GenAI virtual assistant can provide more inspirational bits of advice, guiding the shoppers throughout the whole website navigation journey with contextual and relevant information about the product listed in their vast online catalogue.
In the short term, conversational AI could be an attractive option for retailers in need of powering up their poor e-commerce search experience, as it could be easily done by plugging in GenAI solutions through API rather than reviewing their in-house search capabilities. Ultimately, GenAI could lead to a different kind of product search, giving shoppers the chance of starting a conversation with the retailer/brand by uploading an image, photo, and audio, or by simply speaking to the intelligent AI concierge.
For example, starting from June 8, French grocery retailer Carrefour has integrated a chatbot based on ChatGPT into its company website, giving shoppers the ability to use natural-language AI to assist them in their daily grocery shopping. Customers can find “Intelligent Assistance” on the company home page and consult it in choosing the best products based on their budget, food intolerances and allergies, or even specific menu ideas.
Also, being connected to Carrefour’s website research engine, the chatbot can give a real-time list of products while conversing with customers until the purchase is finalized.
Enhanced Product Visualization Online: Anthropologie, Everlane, H&M and LOFT
Product visualization online has been historically the main source of frustration for shoppers and forgone sales for retailers, especially when the product in question is an experiential one such as a piece of clothing or furniture. Using GenAI for image generation, retailers and brands can optimize the creation of product images, reducing the time and costs associated with in-studio shooting. Indeed, GENAI tools can be used to transform text or 2D images into 3D products representation to offer more appealing and interactive online product catalogues.
H&M, LOFT, Everlane and Anthropologie – among other brands – are all partnering with Google in its latest release of a GenAI virtual try-on tool for women’s apparel on Search. Starting from the brand product catalogue image, the tool can create in real-time an AI-Generated rendering of how it would fit on a diverse set of real models in various poses.
Concerning other previous attempts at virtual try-ons for clothes, such as the AR-powered “Be Your Own Model” released by Walmart, GenAI technology is raising the bar by reflecting accurately how articles of clothing drape, fold, cling, stretch and form wrinkles and shadows- and do it at scale.
GenAI in Retail: What’s Next?
LLMs, FMs and GenAI advancements are happening quickly, especially in comparison to more slow-burn technologies that we have seen emerging in the previous years- such as metaverse, blockchain or AR (Augmented Reality) and VR (Virtual Reality). The immediate applications for the retail industry that we have seen above suggest that retailers and brand executives must start planning their GenAI strategy today, since many of the use cases that might seem so futuristic now could become viable within months rather than years.
In this context, a wait-and-see approach could be risky for retailers. Therefore, they must start experimenting with a test-and-learn approach now to develop a repeatable process that can be deployed tomorrow on a larger scale, while setting the basis in the workforce for GenAI change management and in-house capability building.
