- Amarra is a New Jersey-based manufacturer and wholesaler of special-occasion dresses.
- Over the past five years, it has used AI to help with inventory management and content creation.
- This article is part of “How AI Is Changing Everything: Small Business,” a series exploring how small businesses are using AI for success.
In the ever-changing e-commerce landscape, companies must find a way to be more efficient, competitive, and nimble. And unsurprisingly, some small businesses are turning to AI for help.
Amarra, a company that manufactures and wholesales special-occasion gowns and dresses — including for weddings and quinceañeras — has integrated AI tools into its operations since 2020. Artificial intelligence has helped the company’s leaders better understand customer needs and increase efficiency across such tasks as writing product descriptions and inventory management, Kunal Madan, Amarra’s cofounder, said.
“The primary motivation was to enhance efficiency and elevate the customer experience,” Madan said in an email interview. “E-commerce is a highly competitive field, and leveraging AI allowed us to optimize how we manage inventory, create content, and interact with our customers — which markedly improved our overall performance.”
The company, which is headquartered in New Jersey and wholesales its products to more than 800 independent retailers worldwide, uses AI to optimize its marketing and customer service operations and analyze market trends. Amarra launched in 2018 with just two employees, up to 25 today.
Courtesy of Amarra
Bringing e-commerce and AI together
Amarra uses ChatGPT, a generative-AI assistant, to write product descriptions. The team inputs information on product features and unique selling points into the chatbot, which then produces an initial draft of each blurb. From there, the team refines the text to align with its brand voice.
Its website showcases dresses by color, silhouette, or occasion, each with a multisentence description designed to get a retailer to buy them. Terms for a neon pink ball-gown-style dress include “whimsical” and “fairytale-inspired.” Other blurbs reference floral appliqués and stone accents — all written with the help of ChatGPT.
Amarra’s ChatGPT use, Madan said, has cut down content creation time by 60%, with no loss in quality.
The company also uses an AI-powered inventory management system that predicts stock needs based on historical data and seasonal trends. The tool has reduced overstocking by 40%, Madan said, and has helped Amarra avoid selling out of popular items. This allows the company to ensure product availability for the bridal shops, boutiques, and other independent retailers in North America, Europe, the UK, and Australia that purchase its dresses.
Amarra also uses AI to analyze customer review data across platforms, which helps it better understand shifts in buying behaviors and consumer preferences.
“This helps us identify emerging trends and opportunities for product improvement, keeping us ahead in the market,” Madan said.
AI drives Amarra’s customer service operations, too. The company said that AI chatbots now handle 70% of customer inquiries, answering basic or commonly asked questions and providing faster responses. Madan said this freed up Amarra’s team to focus on more complex issues, such as product customization requests and resolving intricate customer payment issues.
Using AI to complement, not replace
Though AI-driven innovations have allowed Amarra to streamline its operations, adopting the technology hasn’t been entirely smooth.
Madan said one of his team’s biggest challenges with using AI had been “achieving the right balance between automation and the human touch.” For example, Amarra’s chatbot initially provided rigid, impersonal responses to customers.
“We’ve since refined these to better reflect our brand personality, improving the overall customer experience,” he said.
Merging AI systems with Amarra’s existing tools and processes and managing the biases in AI models have been two other challenges. For example, the company’s chatbots initially didn’t account for cultural nuances in different markets, leading to occasional misinterpretations.
It has also seen biases in product recommendations, such as AI favoring certain items without considering newer inventory. Madan said “constant tweaking” helped mitigate these issues. But it’s an ongoing process to address the biases and seamlessly integrate Amarra’s technology environment while keeping data secure.
To ensure responsible AI use, the company actively involves its employees in implementing and customizing the tools. Madan said the collaboration helped Amarra maintain a balance between AI-led automation and human oversight. The company also regularly trains its workers to use AI effectively in their everyday work, he said.
“We focus on using AI to augment our employees’ capabilities, rather than replace their roles,” he said. “AI handles repetitive tasks, allowing our team to invest more time in creative and strategic functions that require a human touch.”
Amarra has learned a lot as it has tested and adapted its AI use over the past half decade. Madan’s advice for other small businesses embarking on a similar journey is to focus on good data management practices since data is the foundation of AI models and algorithms.
They also should consider starting with very specific, manageable areas where AI can have the most impact, such as customer support or inventory management, he added. Any functions should be flexible enough to adapt to changing business and operational demands, while staying true to the company’s ethos.
“Be prepared for a learning curve,” he said. “Some initial hiccups are normal, but iterating based on feedback leads to better outcomes.”