New PolyScore: a simpler, more accurate 1–5 score
New PolyScore: a simpler, more accurate 1–5 score
Rollout date: 5 August 2026.PolyScore — the automated quality score assigned to every conversation with your agent — is moving from a 0–10 scale to a 1–5 scale, backed by a new, fully transparent evaluation rubric.Charts on the Analytics pages and scores on individual conversations in Conversation review will reflect the new score. The new score is calculated for the preceding week.Why we’re making this change
- One score across every channel. The original PolyScore was designed for voice calls. The new rubric natively evaluates voice, messaging, and email — inbound and outbound — with channel-appropriate judgement. For example, a customer silently leaving a webchat after their question is answered is treated as success, not abandonment.
- Simpler and more accurate. The previous score combined five overlapping categories (conversation quality, repetition, frustration, resolution, and task completion). The new rubric condenses these into two scored dimensions plus an engagement gate, each with explicit decision rules — reducing ambiguity and making the score more consistent.
- A familiar scale. The 1–5 scale mirrors CSAT, so PolyScore reads naturally alongside the customer-satisfaction metrics you already use.
As with the previous PolyScore, conversations must meet two criteria to be scored, and both reasons for skipping scoring are now made explicit:
- User engaged — it has to be possible for the agent to do its job. Spam calls, silent calls, or conversations where a user instantly requests a human count as not engaged.
- More than 3 user turns — only conversations with real interaction are scored.
Key rubric decisions
- Handoffs are neutral on task success. From the user’s perspective, being routed to a person is a completed outcome. A handoff is never scored as “not completed” — that rating is reserved for genuine dead-ends where the user got nothing and nobody. A handoff driven by a struggling agent is penalized through the Agent quality sub-score instead.
- Self-service paths count as completed. If the agent gives the user a concrete way to finish the task themselves — for example, “you can reset your PIN any time at acme.com/pin” — that scores the same as resolving it in-conversation.
- Frustration only counts against the agent when the agent caused it. Unhappiness with a policy or outcome doesn’t penalize the agent; being stuck in a loop does.
- Design choices aren’t penalized. If your agent is configured to deflect or decline certain requests, executing that correctly scores as competent handling.
- Outbound declines are neutral on task success. A polite “not interested, remove me,” honored cleanly, is scored as the agent doing its job.