Definition
Bot Performance in Salesforce is a dashboard and reporting feature that provides analytics on how Einstein Bots are performing. It displays metrics such as total sessions, resolution rate, average session duration, transfer-to-agent rate, most-used dialogs, and customer satisfaction scores. These insights help administrators identify bottlenecks, improve dialog flows, and measure the bot's impact on case deflection.
Real-World Example
Consider a scenario where a customer success manager at CloudNine Solutions is working with Bot Performance to streamline support operations and reduce the backlog of unresolved customer issues. With Bot Performance in place, the team routes cases to the best-qualified agents, tracks SLA compliance automatically, and provides self-service options that deflect 30% of incoming volume.
Why Bot Performance Matters
Bot Performance provides analytics and reporting on Einstein Bot activity. Key metrics include total sessions, resolution rate (the percentage of sessions the bot handled without escalating), average session duration, transfer-to-agent rate, most-frequently invoked dialogs, and customer satisfaction scores gathered through post-conversation surveys. Some metrics are aggregated at the bot level, while others can be drilled down to specific dialogs or time periods.
These metrics serve two main purposes: proving the bot's business value and guiding its improvement. A rising resolution rate over time tells you the bot is learning to handle more cases, which translates directly to deflected Cases and staff time saved. A high transfer-to-agent rate for a specific dialog tells you that dialog has a design problem that's confusing users. CSAT scores on bot-handled conversations tell you whether customers are satisfied with the automated experience or merely tolerating it. Mature bot programs review these metrics weekly or monthly to decide where to invest development effort.
How Organizations Use Bot Performance
- •CloudNine Solutions — Reviews Bot Performance metrics weekly during a dedicated bot improvement meeting. Each week they pick the dialog with the highest transfer-to-agent rate and investigate why users are bailing on it, then improve the design in the next iteration.
- •QuickAssist — Tracks bot resolution rate as a key OKR for the support team. When the rate rose from 45% to 62% over six months, the team could directly attribute the improvement to reduced Case volume and used the data to justify continued bot investment.
- •Wanderlust Travel — Uses Bot Performance to compare the impact of dialog redesigns. When they rewrote a booking-modification dialog, CSAT on bot-handled modification sessions rose from 3.2 to 4.1, which confirmed the redesign was worth the effort.