How Nuuly turned AI savings into higher pay for its support team

Nuuly, the clothing rental subscription service owned by URBN, passed $500 million in revenue in 2026, with half a million subscribers. Every one of those subscriptions creates support moments that are personal and time-sensitive, often emotionally charged: a pause, a refund, or a missing item blocking next month's box.

Growing 40% in a single year usually forces a support team to hire, and keep hiring. Natalie Hurst, Nuuly's Sr. Director of Customer Success, rejected that convention. She wanted a team small enough to stay genuinely connected to customers and to each other, whatever size Nuuly reached. Working with Giancarlo Castaldo, AI Content & Performance Manager, she rebuilt the support operation around Fin, the AI Customer Agent. Over three years, she expanded what Fin could do, until it was executing Nuuly's highest-volume workflows end to end.

It paid off. The team has stayed below 80 people, nearly half the size the projections called for. The customer experience Natalie set out to protect stayed strong, with CSAT remaining at 95%. Meanwhile, Fin resolves 35% of the incoming conversations, saving $359,000 in salary costs over six months.

Those savings went back into the team. Nuuly restructured its support team with more career growth and higher pay, moving its most senior associates into high-value retention work. Support now protects revenue rather than just costing money.

Renting clothes creates far more customer contact than selling them. Boxes move in both directions every month, and billing dates matter. Before Fin, anywhere from 30% to 40% of Nuuly's subscribers contacted support every month. The highest-volume requests were pausing, cancelling, or requesting a refund. Each one required an associate to manually cross-check billing status and carrier tracking across multiple systems before they could act.

As subscriber growth accelerated, Natalie ran the math on what keeping pace would take. It worried her. "When I did my five-year plan, my team was supposed to be 120 people by year five. That was super stressful," she recalls. "I've witnessed a number of rounds of layoffs throughout my career, and I'm always conscious of not growing the team too fast and running the risk of needing to let people go."

Her conviction went beyond caution. A small team means she can hire only excellent people and give them time to really know the subscribers they serve. "I think a big contributing factor to our high CSAT scores, which are consistently at 95% or above, is that our support team has a genuine connection to our customers. That's very rare, particularly for the fashion and ecommerce industries," she shares.

The challenge was holding onto that at scale – controlling headcount and protecting the team's culture, while still giving a fast-growing subscriber base instant answers.

Nuuly has run its support on Intercom, the AI-powered helpdesk, for seven years. The team was originally drawn to the way it combined automation with a personal, human approach. When Fin launched, Natalie was one of the first to deploy it. "We wanted to get ahead of AI functionality as quickly as we could and be an early adopter of the technology," she recalls.

But moving fast didn't mean starting over. What appealed to Natalie was layering AI onto the journey her team had designed, not rebuilding around it. "Ultimately, we wanted to maintain the customer journey that we had already created and have Fin be an added layer of efficiency."

To stay ahead, Natalie brought in Giancarlo to own Fin full time. "It was all very new. I didn't totally know what I was getting into, but it became this really integral part of Nuuly and the customer service team," he says.

The rollout started with the foundations every AI agent deployment needs.

1. 1
   The team shaped Fin's voice around the brand's personality and gave it its own name, ChatCat, so subscribers would experience the same warmth from the AI as from Nuuly's people.
2. 2
   From his first week, Giancarlo built out the knowledge base Fin would draw on, so it could answer Nuuly's simplest, highest-volume questions instantly, around the clock. "These agents are only as good as the content you give them. If you can be that librarian for the agent, that sets you up to use the more advanced tools," he says.

The foundations paid off. By the end of Giancarlo's first year, Fin was resolving over 30% of the conversations it took on. A year of accurate answers meant the team was ready to trust it with more.

The second phase connected Fin to Nuuly's own systems, giving it live access to each subscriber's information through Nuuly's API. Answers became personal as Fin could now respond to questions about a subscriber's own billing date, order status, or account. That quickly extended to letting Fin take action on the customer's behalf, which opened up far more complex queries.

The initial test was Nuuly's single highest-volume workflow: pausing subscriptions, cancelling them, and refunding subscription charges when a return was already on its way back.

This process used to take time, both for the associate actioning it and the customer awaiting a response. "It's a lot of clicks, a lot of moving from screen to screen, checking tracking externally, and then coming back to refund if needed," Natalie explains. It happened thousands of times each month.

Now Fin resolves these queries end to end. In about ninety seconds, it triages the conversation, fetches the subscriber's live subscription and order status, and verifies the return against carrier tracking. It then executes the right combination of pause, cancel, and refund before confirming the outcome with the customer.

The team remembers the day it went live. "We had engineers in the room, the product team in the room, and everybody's just watching to see if it would work," Natalie reflects. When it did, the whole team cheered. "The first refund was so fun to see," adds Giancarlo.

Since then, Fin has earned its way into more of Nuuly's complex work. For the most sophisticated workflows, a human-in-the-loop pattern adds control. Take a customer who forgets to include an item in their return. Fin checks their order history, identifies the item, and gathers everything an associate needs before handing over for a quick approval. Nuuly can launch new automations early and keep hold of the sensitive final step.

Extending Fin's capabilities and reach is now part of Giancarlo's day-to-day work. And recently, that work got a whole lot easier.

The newest addition to Natalie's team is Operator, an agent that helps the team run Nuuly's customer operations. Its conversational interface makes it easier than ever to manage the complexity of an AI support org – spotting opportunities in conversations and shipping the improvements.

Giancarlo put it to work right away. "I asked it to point to opportunities for new Procedures based on what we had already built. It immediately suggested one and built it," he says. "A task that probably would have taken a week took 30 minutes. The next day it was resolving at least a third of the conversations it was getting."

Operator also opens up a depth of visibility and insight into the operation that used to be out of reach for Natalie. "I start every Monday asking 'what happened over the weekend?' It reads the interactions and summarizes the volume, the tags, where we saw spikes and why, so I don't have to go through report after report," she says. It helps her field leadership questions too. "If I'm getting questions from my exec team about what we're seeing, I ask Operator, and it puts together the perfect summary."

The result is a team that spends less of its week running the machine and more of it working with customers, which was the goal all along.

Today, around 80,000 conversations come into Nuuly's support team each month, and Fin fully resolves 35% of them end to end, without a human ever getting involved. Over a recent six-month period, Natalie estimates its work equated to $359,000 in salary savings.

However, the number Natalie watches most closely isn't the ROI. "Our team was supposed to be double the size by this time, probably more than double," she notes. This year’s hiring plan required only two new people, while the subscriber base grew well over 30%. And crucially, quality hasn’t suffered, with the customer experience remaining high. "Customer satisfaction is one of our North Star metrics, and our human-AI approach to support has enabled us to maintain a CSAT score of 95% and above," Natalie says.

Staying small has made the team more valuable. With Fin absorbing the routine volume, the team's most experienced people now have time to dedicate to other efforts. "Our senior associates have started doing a lot of retention-based activities, figuring out how to reach people at the exact right moment to prevent their churn," Natalie notes. That work is turning support into a revenue driver in its own right.

"Because of the work Fin is taking on and the way we were able to prove its success, we reorged the team to allow for more career growth and more pay," she says. "The dollars it takes for Fin to do an interaction are significantly less than a human interaction, so I can pay my more senior people way more than I was."

For Natalie, this rewrites an industry default she has pushed against her whole career. "People have always looked at customer service as entry level/temporary roles. It makes no sense to me. It's the most visible role there is. You need high-quality people, and you give them the growth opportunity and the pay to make them stay," she says.

Natalie's targets keep rising. Fin is already involved in around 75% of Nuuly's conversations, and she wants the automation rate at 40% by the end of the year. She is direct about what sets the ceiling, and it sits inside Nuuly rather than inside the AI. "If I had a dedicated person on the engineering team creating API endpoints for us, we would probably have at least a 90% involvement rate and a 60 to 70% resolution rate," she says.

Natalie's three-year plan reshapes the team as Fin takes on more. Today, new hires move from tier one to tier two tickets as they go full time. She wants Fin to absorb tier two questions entirely, so people can jump straight into senior roles. Those include AI specialists, quality experts, and what she calls a "surprise and delight team," welcoming new subscribers and spotting the customers who need a human early in a conversation. The retention mission extends to reaching customers before they ever write in. "I would love for Operator to say, here's a trend I'm seeing, maybe let's do some outreach to these people, and just go and do it," she says.

For Natalie, all of it serves the principle she started with – a small, connected team of people doing the work only people can do, with Fin doing the rest. And for Nuuly's subscribers, the experience stays personal no matter how big the company gets.