~12 agents’ worth of work absorbed in 33 days, and CSAT went up.

A subscription brand whose support load grew faster than the team.

A high-volume subscription skincare brand was answering tens of thousands of customer messages a month, every one of them hand-written. The classic ecom support load was eating the team alive.

• 44,870 messages in a 33-day window, around 1,360 a day, every one written by a human agent.
• Median first response of 1.3 days. Order-status, cancellation, and shipping questions waited behind everything else.
• Repetitive volume crowding out the hard cases. WISMO, cancellations, and troubleshooting consumed the hours that skin-reaction and care cases needed.
• Hiring through the curve was neither scalable nor affordable for a subscription brand riding volume spikes.

They wanted to scale support output, not the headcount.

The brand did not want to keep adding agents to keep up with subscription growth. They wanted the repetitive volume answered autonomously, faster first responses, and quality that held under their own SOPs. The hard requirement: the AI had to keep their care-forward voice on sensitive cases, not read like a script.

Richpanel committed to both at once. Autonomous resolution for the repetitive email volume, and an AI co-pilot drafting in the brand’s voice for the channels the team still reviews. Same team, same SOPs, with the AI doing the typing.

Autonomous on email, co-pilot on the rest, all on their SOPs.

• Autonomous email resolution. 18,074 messages, 83.9% of AI output, sent autonomously on email, with median first response down to 17 seconds.
• Co-pilot on Contact Form and Live Chat. The AI drafts, the team reviews and sends, so quality is protected on the channels they chose to keep in the loop.
• Their SOPs, their voice. No change to the team or the playbook. The AI absorbed the workload of about 12 full-time agents while CSAT on AI-sent messages came in at 4.14 / 5.

Their team types 40% less. Reach grew 8%.

Before AI, every message was hand-written. Now agents type 40% fewer messages while customer reach is up 8%. Effective output is 1.80× their hands alone: for every 100 messages an agent writes, the AI writes another 80.

• AI CSAT of 4.14 / 5 on AI-sent messages, +22% over the 3.40 pre-AI baseline.
• ~12 full-time agents’ worth of work absorbed, 17,908 messages the team didn’t have to write.
• Median first response down to 8.9 hours from 1.3 days, around 3.5× faster overall.
• 21,594 messages sent by AI across 18,777 unique conversations, 44% of total team output.
• Total reach up 8% (44,870 to 48,556 messages) while the team typed 40% less.

33 days from launch to this report.