How Lumi Went from 1 Agent to a 72-person operation – and a 92% CSAT
Lumi Personalized Styling App
Lumi is an AI-powered personal styling app helping over 100,000 women put together outfits every day. The service personalizes suggestions based on style preferences, body type, budget, and lifestyle, then offers to shop the look directly in the app. Whether it's a workday, a weekend, or a special occasion, Lumi takes the guesswork out of getting dressed and helps bring some structure to your wardrobe.
Who’s EverHelp?
EverHelp is an industry-leading provider that combines human + AI efforts to deliver custom support outsourcing services. Since our founding in 2021, we have scaled from 5 employees to 1,000+ agents across 4 continents and have worked with over 100 clients. But great support today isn't just about people, but about pairing the right humans with the right technology.
That's why we combine our expert agents with Evly, our proprietary AI customer service agent built on everything we've learned about delivering exceptional customer experiences. Evly:
- instantly responds to queries in just 15 seconds
- handles 85% of routine tickets without agent involvement
- and reduces workload by up to 60% during peak periods.
The result? A seamlessly blended human + AI support model that helps our clients:
- Deliver quick responses in under 45 seconds
- Offer a true omnichannel experience across all touchpoints
- Achieve 83%+ CSAT scores
- Resolve 97% of disputes
The Challenge: Organizing Support Processes to Match Growing Operations
Lumi's core product is deeply personal, and its users expect real, thoughtful styling guidance. As their support setup wasn't equipped to deliver that, they turned to EverHelp. Prior to the partnership, their customer service:
- Had no organized support processes and no dedicated team
- Relied on 1 agent handling 600 monthly tickets
- Had an FRT of 2 hours and an average resolution time of 4 hours
- Was losing revenue through a 40% refund rate and a 2% chargeback win rate
- Had plateaued at 50% CSAT and a Trustpilot score of 3.0.
Our Solution: Building a Styling-First Support Team for Personalized CX
Though the client initially wanted to build a support system for simple ticket handling, we soon recognized that their audience requires something else. That shaped how we approached Lumi’s case.
Positioning
We analyzed customer insights and helped our client:
- To separate two projects: one focused on the luxury segment (Title), and the other – targeted at the mid-market segment (Lumi)
- To outline Lumi’s positioning as a platform offering AI-powered styling tips and outfits from affordable brands.
Support Operations
We then moved to building a solid support system:
- Established a structured hiring process to bring in talent.
- Developed the full concept and curriculum for an internal training program called Lumi Stylists School to onboard new stylists with the product knowledge and communication skills.
- Sourced a team of 20 support agents to handle email and chat assistance in several languages.
- Introduced a Technical Support agent to the project to handle platform-related technical issues.
- Hired 5 Billing Agents to manage workload related to chargebacks and refunds.
Workflow Optimization
- Built workflows for every support function, from first contact to escalations.
- Streamlined first-contact handling across email and in-app chat.
- Assigned 3 Quality Control agents to monitor the team’s performance and keep the knowledge base up-to-date.
Automation
- Once we established all the necessary processes, we introduced our AI assistant Evly.
- Evly automated repetitive ticket handling, covering 60% of all tickets.
The Results
.webp)
Looking Head
Through our partnership, we managed to provide Lumi with both a support team and a product asset. From here, we are planning to only build on further:
- Systematize Lumi Stylists School to keep onboarding fast and consistent as the team grows.
- Feed stylist interaction data back into Lumi's AI recommendation engine to turn frontline conversations into product intelligence.
- Explore Evly AI for first-contact triage to maintain the 20-second FRT as chat volume scales.
- Maintain curriculum quality and regular stylist coaching to keep CSAT at 92%+ as the business grows.
FAQ
EverHelp brought the chat FRT down to 20 seconds and email FRT to 3 minutes. Monthly request volume grew to 60,000+, and CSAT jumped from 69% to 92%, a 23-point improvement. These results were achieved by combining structured workflows, specialist recruitment, and a purpose-built training program tailored to Lumi's unique product.
The Lumi Stylists School is an internal training program developed by EverHelp from scratch, including its full concept and curriculum. It was created to onboard new stylists with the product knowledge and communication skills needed to deliver consistent, high-quality personalized styling advice. The school ensures every new stylist can hit the ground running without sacrificing the quality users expect.
EverHelp plans to systematize the Lumi Stylists School to keep onboarding fast and consistent as the team scales. They will feed stylist interaction data back into Lumi's AI recommendation engine to turn frontline conversations into product intelligence. Evly AI is also being explored for first-contact triage to maintain the 20-second FRT as chat volume grows, alongside regular stylist coaching to keep CSAT at 92% and above.
Lumi had no organized support processes and no dedicated team in place. A single agent was handling 600 monthly tickets with a first response time of over 20 minutes and average replies taking up to 2 hours. CSAT had plateaued at 69%, well below the standard expected for a personalized styling app serving over 100,000 users daily.
EverHelp established structured workflows and support processes from scratch, built a dedicated hiring process to recruit 36 stylists, and optimized first-contact handling across both email and in-app chat. Every element of the operation was designed with Lumi's personalized user experience at the center, ensuring stylists could deliver meaningful, tailored guidance rather than templated replies.
While Lumi initially approached EverHelp to set up a basic ticket-handling operation, it quickly became clear that their audience expected real, thoughtful styling guidance, not generic responses. EverHelp recognized this and shifted the approach entirely, recruiting 36 specialist stylists who could actively provide outfit recommendations based on each user's preferences, body type, budget, and lifestyle.
More Cases
.webp)
.webp)
.webp)
.webp)
.webp)
.webp)
.webp)
.webp)
.svg)
.webp)
Here’s why clients love EverHelp
agents
time to hire
