In 2022, 38% of retailers reported using AI technology, according to Gartner’s annual survey. This year, that number was close to 50%. By 2025, 94% of retail CIOs think they’ll be using AI/ML. While some large retail brands have been using AI for years, adoption is clearly accelerating across the industry. The launch of ChatGPT supercharged interest in generative AI (GenAI) in particular, and retail leaders are eager to see how GenAI can drive value for their businesses.
As GenAI tools continue to improve their ability to mimic human language and engage effectively with humans, the most obvious GenAI use cases in the retail industry are customer support and content creation for products and marketing. For retailers, customer experience is everything. If new GenAI capabilities can bring increased precisions, personalization, and cost-saving automation to customer engagement, it would be a huge win.
Even so, retail leaders need to start eyeing other GenAI use cases. Yes, GenAI will likely revolutionize customer engagement in the next several years. But if they stop at chatbots, virtual assistants, product descriptions, and agent support tools, retailers will miss out on some of the deeper, more fundamental impacts than GenAI can have on their businesses.
Beyond Customer Engagement and Software
The buzz surrounding ChatGPT has naturally highlighted GenAI’s text-creation capabilities. Retailers are also excited about the role GenAI will play in the software development lifecycle (SDLC). But GenAI can also make an impact in less obvious ways, including:
- Merchandising
Retail begins with the art of identifying and sourcing the right product, the right range, and the right price. GenAI can be the merchandiser’s sidekick, synthesizing data across customer search patterns, social media, and product attributes to optimize the performance of the product mix. Leveraging both historical data and emerging trends, GenAI may be able to enhance the pricing optimization capabilities (and will certainly enhance the user-friendliness) of current AI-driven pricing tools, which could reduce the need for discounts while increasing margins and shelf throughput. - Store Operations
Yes, GenAI will certainly play a role in the customer engagement piece of store operations, providing next-best-actions to store associates and even enabling virtual shopping assistant avatars for fashion brands. But GenAI will also be able to guide stock recommendations and operational decisions for individual stores based on footfall patterns, weather, seasonal buying patterns, local events, and other factors.
- Supply Chain
GenAI can suggest supply chain performance improvements related to demand planning, transportation routes, delivery/fulfillment methods, and warehouse space. It can also support supplier relationships by helping negotiate better prices. Furthermore, retailers will leverage GenAI to optimize sourcing by providing micro-detailing inputs such as SKU-level demand, individual supplier performance, and real-time market data.
- Digital Channels
GenAI will soon be generating significant amounts of content for digital channels: product descriptions, product images, reviews, product guides, and even product videos. But behind the scenes it will also go far beyond creating consumer-facing written content and images. It will optimize SEO content, create webpage designs, and suggest real-time online pricing adjustments.
None of this is to say that retailers should de-emphasize the more obvious sales/marketing and SDLC capabilities of GenAI. However, as they become more comfortable with GenAI, retailers should aggressively seek to deploy it in other ways that add value.
A GenAI Roadmap for Retail
All AI starts with data. In a GenAI world, building algorithms becomes less important while data becomes even more important. The leading large language models (LLMs) will suffice for most retailers, but to deliver value, the models will need access to a robust, modernized data estate.
Many retailers are already struggling to synthesize and act on all the structured data streaming in from purchasing systems, POS terminals, and eCommerce channels. To deliver value, GenAI tools will need to consider a much wider net of unstructured data, including social media feeds, online product reviews and ratings, helpdesk emails, SOPs, and quality documents. To build a viable GenAI strategy, retail technology and business leaders first need to solve the problem of overlaying GenAI on top of these legacy technologies and data architectures to achieve coherence. Often, this will mean relying much more heavily on cloud providers for both data storage and computing power.
Retailers also need a strategy layer for GenAI. Increasingly, GenAI itself will be able to issue helpful strategic guidance for senior leaders, but business strategy isn’t going to be fully automated any time soon. Do retailers want GenAI outputs that will satisfy their current customer base, or do they want to expand or shift their customer demographics in particular directions? The impact of GenAI initiatives will be only as good as the overall retail strategy that drives them.
Success will require a clear-eyed analysis of where and how GenAI can drive brand value. As they explore GenAI use cases, retailers need to:
- Assess the maturity of systems and processes, particularly to determine the organization’s current capabilities related to data-driven autonomous decision-making
- Identify where it’s more practical to use human intelligence than machine intelligence
- Develop a POV on how the brand plans to use AI, then align senior executives on that strategy
- Build a strong data model, as even the most powerful GenAI tools are next-to-useless without a rich data universe
- Focus on specific KPIs/functions and take an agile, test-and-learn approach to ascertain where GenAI can drive the most business value
- Effectively train employees to co-work with GenAI
