Customer Service That Doesn't Scale Linearly
Customer inquiries scale with revenue. That's just how it works. Sell twice as much, field twice as many questions about sizing, shipping, returns.
The traditional answer is hiring more support staff. More tickets means more people. Linear scaling. It's expensive and it creates its own management overhead.
Little Lies solved this differently. They use a Gorgias AI chatbot (they call it Stevie) to handle the bulk of inquiries, and one person (Michelle) to handle everything Stevie can't. The cost structure is basically fixed regardless of volume.
It's not revolutionary technology. It's just using platforms as intended instead of defaulting to hiring.
Where the cost comes from
At 30 people, Little Lies had multiple staff on customer service. Answering emails, processing returns, tracking shipments, explaining policies. Standard operation for any brand doing volume.
The problem isn't the work itself - most inquiries are straightforward. Where's my order? What's your return window? How does sizing run? Questions that don't require creativity or judgment, just accurate information delivered quickly.
The problem is that human CS teams create variability. People get tired, interpret policies differently, have bad days, need training on updates. And they cost roughly the same whether they're handling 50 tickets or 500.
When Jade, its founder, rebuilt the business, she looked at customer service not as a people problem but as an information delivery problem. Most queries don't need empathy or complex reasoning. They need speed and accuracy.
How “Stevie” actually works
The chatbot integrates directly with Shopify. Customers asks where their order is, it pulls tracking info and responds in seconds. Return request? It walks them through the process and initiates it. Sizing question? References product data and historical feedback.
Jade describes Stevie with actual affection. It's been trained on brand voice (specifically how Michelle talks to customers), handles edge cases reasonably well, and occasionally produces responses better than a human would write.
Stevie's part of the team, she says. They monitor conversations, refine responses when needed, but mostly it just runs. 24/7, instant responses, perfectly consistent.
Michelle handles everything that needs actual judgment - complaints, complex situations, VIP customers. She's not buried in basic tickets, so she can actually solve problems.
The customer experience side
You'd expect customers to hate this. Automated responses, no human touch, chatbot friction.
Turns out the opposite is true. For simple queries, customers want answers immediately. They don't want to write an email and wait 24 hours to find out where their package is. They want to know now, while they're thinking about it.
The chatbot gives them that. And when they do need human help, they get Michelle - who has time to actually engage because she's not drowning in basic inquiries.
We've had customers comment on how much better our service is, Jade notes. Because it is. The bot is instant, and the team isn't stretched thin anymore.
Implementation reality
This only works if you do it properly. The chatbot needs access to real-time data (order status, inventory, shipping), comprehensive policy information, and ongoing training on how to handle different scenarios.
You can't just flip it on and expect it to work. Little Lies monitors conversations regularly, identifies where Stevie struggles, refines responses, updates training. It's maintenance, but it's manageable maintenance for one person, not the ongoing overhead of managing a CS team.
And you need clear handoff protocols. When does Stevie escalate to Michelle? How do you flag urgent situations? What queries should never be automated? Those decisions matter.
Why this matters for growth
The economic shift is significant. Customer service used to scale linearly with revenue - more sales meant proportionally more CS costs. Now it's mostly fixed.
If Little Lies doubles revenue, Stevie handles it without breaking. Michelle might need backup eventually, but not immediately, and not at the same rate.
That's the difference between profitable growth and growth that just creates more work at the same margins.
For any brand doing volume: if you're still paying people to answer where's my order fifty times a day, you're leaving money on the table. Those are platform problems, not people problems.
This article is part of our Open Tabs series, where we talk to founders about the realities of running an ecommerce business day to day. Watch the full episode with Jade from Little Lies here.
