How Iteration Boosted a Support Bot to 70% Resolution (Full Transcript)

A support agent improved from ~10–15% to 70% resolution after changes, driving longer conversations and strong user preference and gratitude.

Download Transcript (DOCX)

Planning vs. Shipping in AI Deployment

Added on 04/16/2026

Speakers

Add new speaker

[00:00:00] Speaker 1: The first iteration of that had like a 15%, I actually think less, like a 10% resolution rate, and then it would kick out to our actual support team to go help fix that thing. And then we got in there and we made a bunch of changes, and now it's up to 70% resolution rate. And it's so good that people are having really long conversations with it and writing in like, wow, that was such a good... That was so helpful. Thank you. Now they prefer talking to that agent. They really like it.

Summary

Speaker describes improving an automated support agent: initial resolution rate around 10–15% before escalating to human support; after changes it reached ~70%. Users now have longer, positive conversations and prefer the agent, expressing gratitude and finding it very helpful.

Copy

Download

Title

Support Bot Resolution Rate Improved to 70%

Copy

Download

Keywords

support automation Remove

Remove

chatbot

Remove

resolution rate Remove

Remove

customer support Remove

Remove

iteration

Remove

escalation Remove

Remove

user feedback Remove

Remove

conversation length Remove

Remove

agent improvement Remove

Remove

Copy

Download

Key Takeaways

Early versions of automated support may resolve only a small share of issues and require human escalation.
Iterative improvements can dramatically raise resolution rates (to ~70% here).
Higher effectiveness leads to longer user engagement and positive feedback.
Successful support agents can become the preferred channel over human support for many users.

Copy

Download

Sentiments

Positive: Tone is upbeat and proud, highlighting strong performance gains (10–15% to 70%) and enthusiastic customer feedback like 'so helpful' and preference for the agent.

Copy

Download

Enter your query

{{ secondsToHumanTime(time) }}

Back

Forward

{{ Math.round(speed * 100) / 100 }}x

{{ secondsToHumanTime(duration) }}

Select Audio file