Member Exclusive   //   April 13, 2023

Unpacked: From marketing to personalization, what generative AI means for retail

As AI-friendly tools grow in popularity in recent months, generative AI has been generating buzz in the retail industry.

Generative artificial intelligence, at its most basic level, is a form of machine learning that can produce new content based on past experiences — be it text, image and computer code, among others. ChatGPT is a generative AI application that’s gaining plenty of attention at the moment and is a chatbot from Microsoft-backed startup OpenAI. What sets generative AI apart is its ability to form human-like responses and learn from interactions.

Generative AI tools are all the rage lately: from students using them to write essays to tech personalities like Elon Musk urging a pause in AI development, the technology has been dominating headlines over the last couple of months. This technology has the potential to impact various areas in retail — including marketing, merchandising and customer service. In the long run, industry experts expect generative AI to boost both productivity and creativity within an organization. However, generative AI is already experiencing some roadblocks with various governments exploring ways to regulate it.

“The business community got excited like any new tool or shiny object that comes out in the market,” said Bharath Thota, partner and chief solutions officer in the analytics practice of global strategy and management consulting firm Kearney. “They were excited with the potential of ChatGPT. At the same time, they were scared off like what it can do to their business.”

Generative AI’s potential in retail

Retailers are still in the early stages of applying AI in their operations. 

At the moment, a number of brands have begun using this technology for marketing purposes. For example, Coca-Cola released an AI platform that combines OpenAI’s GPT-4 and DALL-E in partnership with OpenAI and Bain & Company. The beverage giant encouraged people to create art through the new platform for a chance to be featured on its digital billboards in New York City’s Time Square and London’s Piccadilly Circus. Meanwhile, Patrón Tequila released an AI art generator that creates personalized images of its margaritas.

Barry Thomas, senior thought leader at Kantar, said that retailers and third-party sellers can also utilize generative AI to create better product descriptions. Helping vendors and merchants develop better product descriptions is exactly the idea behind Shopify’s new AI tool introduced last month. 

Thomas said that retailers with more proprietary data have a greater competitive advantage than others. He added that this could push even more retailers to double down on their attempts to acquire even more first-party or zero-party data. “You’ve got so many more possibilities for creation based on the larger data set, the better quality data set,” he said.  

Another use-case is on products recommendation. While retailers haven’t adapted generative AI for this purpose yet, other industries have begun to test out generative AI’s recommendation abilities. Full-service travel brand Expedia said it is beta testing ChatGPT in its mobile app to give users personalized travel recommendations and planning assistance.

Experts believe more use cases for generative AI will appear. Funding for startups driving the growth of generative AI and other AI tools hasn’t slowed despite the tough venture capital environment for everyone else, said XRC Ventures General Partner Al Sambar. Sambar said that the amount of AI-related startups XRC is invested in is estimated to approach 50%. 

Indeed, startups working on generative AI tools are witnessing a gold rush. For instance, Mobius AI, which is founded by former AI researchers at Google, was valued at $100 million just a week after it launched. Founded by former employees of OpenAI, the start-up Perplexity AI recently raised $26 million in funding.

“We see really active investing with generative AI [or] AI in general,” Sambar said. “There are a lot of investors and they’re willing to write capital to companies that are still really early and maybe not yet EBITDA positive.” 

Every now and then, certain startup categories will gain a vast amount of funding but that doesn’t always signal long-term success. Web3 caused a buzz in the VC community too, but funding for Web3-based startups in the third quarter of last year dropped to its lowest since end of year 2020, according to Crunchbase data. Just a few years ago, investors also rushed to buy now, pay later players yet some of the largest companies in the category still aren’t profitable at the moment.

International criticism

Generative AI’s fame comes with criticism. Countries are mulling over the appropriate response to a technology that’s brand new and quickly growing in popularity. Its long-term impact remains unknown, which experts say can be worrisome for regulatory bodies. These obstacles could prevent generative ai from taking off and pose limitations for retailers that invest in it. 

The rules around generative AI are still being written. Earlier this month, Italy became the first Western country to ban ChatGPT due to privacy concerns. The Italian Data Protection Watchdog told OpenAI to temporarily halt processing Italians’ data. 

As opposed to developing new regulations, the U.K. government, on the other hand, told various regulators to devise “tailored, context-specific approaches that suit the way AI is actually being used in their sectors.”

“This is now like building a plane as we are trying to fly it,” Kearney’s Thota said. “You have to be very careful about where you’re investing, how much you’re investing and how things are shaping up in the market.”  

In the U.S., the Biden administration has just begun looking into possible rules for AI tools like ChatGPT. The Commerce Department this week released a public request for comment on possible accountability measures for AI models. The comments will be accepted within a span of 60 days and would help guide lawmakers on how to handle AI.

Want to discuss this with our editors and members? Join here, or log in here if you're already a member.