Building with words: Retail storytelling with WRITER’s Ranjan Roy

Picture this — it’s 1983. Ronald Reagan is president, MTV still plays music videos, and somewhere in America, a father brings home something that’ll change everything for his young son. Not an Atari. Not a Walkman. An IBM personal computer.

That boy was Ranjan Roy, our latest guest on Humans of AI. Today, he serves as the retail lead at WRITER. But Ranjan’s path to becoming an AI pioneer wasn’t a straight line. He shares his story of how he went from being a “policy debate nerd” to a pioneer in bringing generative AI to retail operations, and how he’s using his expertise to help some of the world’s most recognizable brands create more personalized, more human, and more memorable customer experiences.

Tracking patterns from politics to markets

While that early computer sparked Ranjan’s love of technology, high school took him in an unexpected direction — policy debate. For those unfamiliar with this intellectual sport, imagine two teams armed with towers of evidence cards, speaking at breakneck speed about complex governmental policies, all while making split-second strategic decisions.

“So actually not tech nerd, [but] a policy nerd reading about congressional legislation and things like that,” Ranjan recalls. Policy debate taught him to take vast amounts of information and make sense of it under pressure. This skill would be the thread connecting every chapter of his professional life.

Ranjan’s early career reads like someone following breadcrumbs of data across industries. After college, he landed at Bank of America’s emerging markets division, where he spent seven years monitoring multiple screens tracking financial flows.

When the 2008 financial crisis hit, Ranjan made his next move — business school, followed by a stint at the Financial Times. When he arrived in 2011, he was at the center of media’s digital transformation.

“That’s when I think started to get at least a bit of the product bug. And this is 2011 to 13, so social media is starting to pick up,” he recalls. “The way we get our information is changing.”

Following data to the retail industry

Ready for a plot twist? Ranjan joined Adore Me, a direct-to-consumer (D2C) lingerie brand, as their growth strategist. If you’re wondering what a policy debate champion turned finance guy was doing in women’s intimate apparel, so was Ranjan initially.

But Adore Me taught Ranjan that retail isn’t really about selling products. It’s about selling stories. And stories are made of words. And words, as it turns out, are data.

“I started there in late 2018 and watched the company go from 70 million to 300 million plus in revenue,” Ranjan explains.

During a pandemic, while every other retailer was panicking about supply chains and customer behavior, Adore Me was quietly becoming a case study in how to read the room. And Ranjan wasn’t just watching this growth — he was helping engineer it.

“I’ve been told one of the things I’m good at is connecting dots that are sometimes unexpected,” Ranjan says. “I think it’s having kind of a diversity of experience, but also like throughout it all, it was kind of taking large amounts of information and trying to make sense of it.”

Tackling the content creation nightmare

By 2022, Ranjan faced a challenge that millions of ecommerce professionals know well — the content creation bottleneck. Adore Me needed hundreds of unique product descriptions monthly, each requiring 40+ hours of creative work. The team was spending entire work weeks every month trying to find fresh ways to describe fundamentally similar products.

Most companies accept this as an unavoidable cost of doing business. Ranjan saw it differently. He’d been experimenting with early AI tools, and Adore Me was already using WRITER for style guidelines. The most fascinating part is that Ranjan didn’t discover AI because he was looking for the next big thing. He discovered it because he was looking to solve a very specific, boring problem.

“We were able to automate all of our product description creation before ChatGPT even launched in the summer of 2022,” Ranjan says. “It was at a time when no one cared about this, which was kind of fun. There was no hype. It was just simply, here’s an interesting technology. Can we do something useful with it and actually make it work?”

Turning boring tasks into building blocks

Here’s where Ranjan’s story becomes everyone’s story. Because AI’s real power isn’t replacing human creativity — it’s eliminating the mundane work that prevents human creativity. When copywriters stopped spending weeks on basic product descriptions, they could focus on strategy, brand messaging, and customer experience innovation.

“Starting with work that people did not want to do made it even better,” he says. “Because then suddenly it’s like, ‘Oh wait. That kind of work we can start using technology to solve, so then I can actually work on fun, more creative stuff.’”

The initial use case of product descriptions became building blocks for something larger. Ranjan’s team began creating agentic workflows. It sounds fancy, but it’s just AI systems that can chain together multiple tasks. Each output becomes input for the next process.

A single product description could spawn versions for Amazon, Walmart, Target, and Macy’s — each optimized for different platform requirements. Add customer purchase history, product reviews, and behavioral data, and suddenly, you’re orchestrating personalized customer experiences at scale.

Realizing language is the interface

This is around when Ranjan realized something profound about AI’s true nature. You’d think it’d be about ‌technology, but it’s‌ about language. Ranjan spent most of his time — not on technical implementation — but on communication and language. Or words. The thing he’d been working with since high school debate. The thing that connects policy analysis, and financial markets, and retail storytelling.

“To me that’s why large language models and generative AI is actually even more interesting and applicable to me because using words to get a desired output is basically what you’re doing,” he explains. “And I think that’s what makes it different than other types of, certainly coding, but also just technology as well. It’s more kind of like humanities than computer science or the science side of it.”

For the first time in the history of computing, the people who are best with language — writers, communicators, storytellers — can also be builders.

“It allows you to build things that never would’ve been possible in any other context,” Ranjan explains. “You would not have had to either been able to develop that skill, had the time to develop that skill, and now you can just hack things together, build things, have fun, do creative things in different ways using language.”

Evolving from builder to teacher

Ranjan’s transition from builder to teacher happened naturally when WRITER recognized his success at Adore Me and brought him on to share those insights with other companies. But as he began having conversations with other retail brands, he discovered that most of their initial AI initiatives had previously failed. Not because the technology was inadequate, but because organizations had approached it wrong.

According to WRITER’s 2025 enterprise AI adoption report, 68% of C-suite executives report generative AI, creating tension between IT teams and other business areas. 72% say their company develops AI applications in a silo. The difference between success and failure often comes down to who controls the AI interaction.

“We saw in product descriptions actually having a copywriter own the prompt layer changes the entire thing,” Ranjan says. “Because then they start to understand the value that they can create. They start to see how it all works.”

When IT departments manage AI as a black box, adoption often stalls. The best AI implementation requires business and IT departments working together.

The crawl-walk-run philosophy

Ranjan’s approach to AI implementation challenges the “transformation” narrative that dominates business media. Most companies want to rewire their entire organization on day one. Instead of revolutionary change, he advocates for evolutionary progress — crawl, walk, run.

Crawling means starting with specific, measurable problems that AI can clearly solve — like those product descriptions. Walking means expanding successful approaches to related challenges. Once Ranjan had that foundation in place, additional use cases like translating product detail pages became surprisingly quick and straightforward to implement.

“Once you kind of get that core infrastructure built on relatively straightforward problems or your easier ones, you can start to tackle the more complex problems,” he explains.

Running means building AI-powered capabilities that would have been impossible before. That’s when you can start envisioning those transformative possibilities.

Imagining tomorrow’s shopping experience

Ranjan sees a future of hyper-personalization where technology understands not just what customers bought but how they talk about their purchases, when they prefer to shop, and how they like to discover new products.

He thinks commodity purchases will increasingly happen through conversational interfaces — voice commands for toilet paper, chatbots for household supplies. But discovery and browsing will evolve in more advanced directions.

“There’s still, especially in retail, a strong value to overall just the experience of discovery and shopping, and I don’t think that goes away in any way because that’s an innate human trait,” Ranjan explains. “I think there’s gonna be like a lot of innovation around what it means to experience that discovery and the just browsing shopping side of it.”

In the end, Ranjan’s story is not just about AI or retail — it’s about what it means to be human. As we hurtle headlong into a future where machines are increasingly capable of performing tasks that were once the domain of humans, it’s easy to get caught up in the hype and forget the most important thing — people.

But as Ranjan’s journey shows, the future of work isn’t a zero-sum game between humans and machines. And it’s curious professionals — not just computer scientists — who see problems worth solving and aren’t afraid to experiment with new tools that are building this future.

Want to hear more stories from the humans working at the crossroads of business and generative AI? Subscribe to Humans of AI wherever you listen to podcasts.