Product updates
Kustomer introduces shared QA scorecards for human and AI support
Vendor-reportedChecked: [qa]
Editorial attribution: Gosia Madej
Prepared with AI assistance. Editorial attribution does not imply personal product testing or manual review of every story.
Published: · Event date:
Kustomer’s October 14 announcements add automatic conversation fields and a common QA rubric for human and AI agents. Conversation Intelligence is listed as available; customers are directed to confirm their AI QA rollout timeline.
Vendor-reportedChecked: [intelligence][qa]
What changes for users
Before using an aggregate QA score for coaching, evaluate a closed handoff in which the AI gives a wrong answer and a person corrects it. Check whether the explanation separates the original mistake, the handoff and the recovery. Keep an AI-inferred satisfaction field distinct from customer-submitted feedback, and verify that a manual correction survives reevaluation.
Editorial assessmentChecked: [intelligence][qa]
Conversation Intelligence classifies requests, estimates scores and extracts details into normal Kustomer fields. The vendor says those fields include model reasoning and can drive reports and workflows. Its satisfaction score is inferred from conversation content, rather than collected from a customer survey.
Vendor-reportedChecked: [intelligence]
AI QA applies a weighted scorecard to closed conversations handled by people or AI agents. For a conversation that passes from an AI agent to a person, Kustomer describes one assessment of the full exchange. Admins can configure evaluation criteria and timing, and manually corrected values are preserved on subsequent runs, according to the vendor.
Vendor-reportedChecked: [qa]
A shared rubric does not establish responsibility for each participant’s actions. The launch account describes end-to-end handoff assessment, but does not demonstrate reliable attribution when an AI error is later repaired by a person. Broad scoring coverage and correct evaluation remain separate questions.
Editorial assessmentChecked: [qa]
Checked against sources
AI-assisted source check
How this was checked
AI-assisted checks read both complete October 14 Kustomer announcement blogs and the linked product pages. These are vendor sources, not independent evaluations. The separate press-release page remained inaccessible; no product test or source screenshot was obtained.
Editorial assessmentChecked: [intelligence][qa]
Conversation Intelligence is described as available now. The AI QA post asks existing customers to discuss setup and rollout timing with their Customer Success Manager; it does not establish access for every account. We could not verify numerical pricing, regional eligibility or scoring accuracy from these announcements.
Editorial assessmentChecked: [intelligence][qa]
Source evidence
Original source excerpt. This shows what the source says; it does not establish product performance.
Once corrected, that value won't be silently overwritten
Sources
- [intelligence] Kustomer: Introducing Conversation IntelligenceVendor-reportedPublished: 14 October 2026 · Checked: 15 October 2026
- [qa] Kustomer: Introducing AI QA for human and AI conversationsVendor-reportedPublished: 14 October 2026 · Checked: 15 October 2026