There’s no hard and fast rule for when to invest in new technologies, said David Cost, Chief Digital Officer at Rainbow Shops on a recent “Reimagining Retail” episode.

“We don’t know [when the right time is],” he said. "We take bets, and we try to take bets that have limited downside if we're wrong."

Rainbow operates over 850 stores across the United States and US territories, selling affordable, size-inclusive clothing for women and kids primarily in urban and multiethnic neighborhoods.

The retailer has spent years investing in digital transformation while serving a highly value-conscious customer base, and the lessons learned challenge many assumptions about technology adoption in retail.

Waiting for mainstream AI adoption

Despite the hype around AI in retail, Rainbow's data shows limited adoption among its customer base. The company launched a post-purchase survey asking customers if they use ChatGPT, and results have held steady at about 25%, meaning three-quarters of customers don't use large language models (LLMs).

"In our world, when we go to ecommerce events, you would swear absolutely everyone spends all their life in [ChatGPT] or Claude or Gemini or whatever the tool of their choice is," Cost said. "We just don't see that in the population overall."

Cost's biggest focus for the future is watching when AI usage crosses over to mainstream adoption.

"Today, we're talking about the upper 25%," he said. "It's not a mainstream technology yet. When it hits that point, when we start to see that crossover, then I think things change."

Generative AI's biggest retail impact will be in imaging

Rather than agentic commerce or chatbots, Cost believes genAI will transform ecommerce through product photography.

"In the ecommerce world, we sell pictures, because people can't touch and feel the goods," he said. "Product photography is really expensive and hard."

Rainbow tested this hypothesis with email marketing. The company took a single email and used AI to generate variations of the hero image while keeping everything else (subject line, text, click-throughs) identical.

The result was “material differences” in performance from one image to another, Cost said.

Rainbow already uses genAI for writing product titles and descriptions, and increasingly for generating product photography. The technology enables unlimited variations for A/B testing at minimal cost.

Looking ahead, Cost envisions personalized product images tailored to individual shoppers.

"An image that works for you may be different than an image that works for [someone else], and so now we're gonna do personalized images for each of you when you get to a website," he said.

Why agentic commerce faces major barriers

The retail industry's enthusiasm for agentic commerce, AI agents that shop autonomously on behalf of consumers, overlooks fundamental behavioral and economic realities, Cost argued.

"Let's say you've had a person's been cleaning your house for the last 10 years," he said. "When they need cleaning supplies, do you hand them your credit card and let them go to the store and buy them? Or do you take the list and go buy the stuff yourself?"

The trust barrier is even higher for Rainbow's customers, many of whom rely on debit cards as their only form of electronic payment. When a debit card is used online, it places a hold on funds, freezing money that customers may need for rent or other essentials.

"No one with a debit card is gonna give that to an agent to go shop on their behalf, 'cause if it makes a mistake, then they can't pay rent," Cost said.

Additionally, polling shows the general public "is angry and distrusts AI," Cost noted. "I just don't see behaviorally how we're getting to the point that we're gonna hand payment details to something that we don't really know and understand."

Retailers should temper expectations around autonomous shopping agents and focus on AI applications that assist rather than replace human decision-making. The technology needs to become mainstream and trusted before consumers will delegate purchasing authority.

Listen to the full episode

We prepared this article with the assistance of generative AI tools and stand behind its accuracy, quality, and originality.

This was originally featured in the Retail Daily newsletter. For more retail insights, statistics, and trends, subscribe here.