How Tremendous reached near-90% support deflection with Decagon across chat, email and voice
Tremendous, which sends rewards and payouts, replaced fragmented chat and email automation and a growing outsourced support team with Decagon across chat, email and voice. As reported by Decagon, deflection is near 90% and CSAT reached 92%.
The problem
Tremendous's weekly support volume grew from roughly 2,500 tickets to nearly 60,000 interactions over four and a half years, and most of it came from reward recipients. Scaling an outsourced support team of about 18 agents was hard. Separate tools handled chat and email, and humans still had to deal with any request that needed more than a knowledge-based answer.
How they did it
- Started with chat and email, the highest-volume channels, then extended to voice.
- Had an initial implementation in place within 30 days that matched the previous tools.
- Built Agent Operating Procedures (AOPs) for high-volume requests that needed a human, starting with reward resends verified through two-factor authentication.
- Added AOPs for cancelling payouts and exchanging rewards, and custom AOPs per enterprise client with different automation and escalation levels.
- Trained support staff to write and debug AOPs, use Duet to analyze conversations, and use Autopilot to find automation opportunities.
- Named an AI auditing and automation lead to maintain agent performance.
Results
- As reported by Decagon, recipient support deflection is nearly 90%, compared with a peak of about 65% on chat with the previous tool.
- As reported by Decagon, CSAT was 92% over 12 months, compared with 86% for the outsourced team.
- As reported by Decagon, first response time went from about one business day to immediate, and average resolution time from about one business day to 3.5 business hours.
- As reported by Decagon, an internal tool for managing abusive recipients was built with Duet in a matter of weeks, and trend questions that took manual searching now take about five minutes.
As reported by the source (Decagon customer story); AgentGid did not measure these figures.
This fits a support team already handling tens of thousands of interactions a week, with someone able to write and maintain action workflows like the AOPs and audit the agent's output. Decagon has no public pricing, so budget needs a sales conversation first, and every result here is vendor-reported. For a smaller team, Rasa (Free) is open source, but its free license caps at one bot and 1,000 external conversations a month.
The agent used here
Similar use cases
How Mercado Libre replaced its voice IVR with Decagon AI agents across LATAM
Mercado Libre, a Latin American ecommerce and fintech platform, replaced its rigid voice IVR with Decagon's voice agents. The team gained natural conversations, no-code control ov…
The team can now see why customers reach out, an insight the article says did not exist before.
How Duolingo's English Test team automated chat support with Decagon
The Duolingo English Test support team replaced its previous AI vendor with Decagon to handle live chat inquiries. According to Decagon's customer story, the agent went live withi…
As reported by Decagon: 80% chat deflection from the start, with a majority of chat inquiries fully resolved.
How Chime scaled chat and voice member support with Decagon
Chime, a US fintech, picked Decagon to automate member support on both chat and voice after comparing several vendors. The company reports high resolution rates, lower support cos…
As reported by Decagon: 70% chat and voice resolution, a 30 sNPS, and a 60% decrease in customer support costs.