Skip to content
ai-agent.review — AI Agent Reviews, home
← All news

Product updates

Parloa announces Relationship Score with a ceiling for trust failures

Vendor-reportedChecked: [release]

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:

The composite score puts conversation experience ahead of efficiency. A capped rating records a detected problem; it does not establish that the agent prevented it.

Editorial assessmentChecked: [release]

What changes for users

Before using the score as an executive KPI, compare its component judgments with reviewed support cases and customer feedback. Ask for weights and cap thresholds, and track detected trust failures separately from the overall trend.

Editorial assessmentChecked: [release]

Conversation and agent-quality signals feed a gauge, while a separate trust check limits its upper range.
Original conceptual illustration of score weighting and a trust ceiling; not a Parloa interface or measured result. · AI Agent Reviews

Parloa announced Relationship Score on October 8. It combines LLM-judged conversation signals into a weighted score out of 100. Conversation experience carries the greatest weight, agent quality less, while trust and compliance failures cap the result.

Vendor-reportedChecked: [release]

Parloa lists personal-information leaks, hallucinations, toxic content, scope violations and missed disclosures as reasons for the cap. The announcement offers factor-level drill-downs but does not give numerical weights or cap thresholds.

Vendor-reportedChecked: [release]

The underlying Lens product predates this announcement: Parloa described it on June 22. That source separates operational signals, such as tool-call errors, from automated behavioral evaluations and describes exports to external analytics tools.

Vendor-reportedChecked: [lens]

Checked against sources

AI-assisted source check

How this was checked

AI-assisted checking read the announcement, release index and earlier Lens description. These are vendor sources; no product testing, independent performance validation or human review occurred. Source screenshots could not be captured.

Editorial assessmentChecked: [release][releases][lens]

The announcement does not specify plan eligibility, activation timing or pricing. It does not establish how accurately the score predicts loyalty or customer satisfaction.

Editorial assessmentChecked: [release]

Source evidence

Original source excerpt. This shows what the source says; it does not establish product performance.

Trust and compliance set a ceiling.

Read original source [release]

Sources