Start Small: First 7 Days
This page is a conceptual rollout plan designed for new customers and evaluators. It avoids UI details so it can be used as a quick “what success looks like” guide.
For step-by-step setup, see the Conversation Analytics – Administration Guide (tenant admins) or Platform Setup & Operations Guide (operators).
Day 1–2: Confirm data readiness
- Ensure your environment is receiving conversation records (calls, and/or enabled text channels).
- For voice calls, confirm transcription is enabled and transcripts look reasonable.
- Identify 2–3 teams/queues to pilot first.
Success criteria - You can open a conversation and see a transcript/thread and core metadata.
Day 2–3: Enable a small set of insights
Start with a “starter pack” that provides fast value:
- Conversation summarization
- Sentiment
- One metric (CSAT is a common starting point)
Success criteria - Insights begin to appear for pilot conversations (not necessarily all conversations—filters may apply).
Day 3–4: Validate and calibrate
- Review a sample of conversations across pilot teams.
- Check both values and explanations.
- Note edge cases and clarify definitions (what “resolved” means, what “dissatisfied” looks like in your org).
Success criteria - Stakeholders agree outputs are directionally correct and explanations are helpful.
Day 4–6: Roll out dashboards and saved searches
- Create dashboards for the pilot insights (e.g., CSAT distribution and trend).
- Create saved searches for operational use:
- “CSAT < 3”
- “Negative sentiment”
- “Escalation reason = …” (if enabled)
Success criteria - Supervisors can drill down from a dashboard bucket to the underlying conversations.
Day 6–7: Expand and operationalize
- Expand to additional teams/queues.
- Add 1–2 additional insights based on what the pilot revealed.
- Define a weekly review cadence:
- top drivers of dissatisfaction
- trending topics
- coaching themes
Success criteria - Conversation Analytics becomes part of weekly operational routines.