Agent categoriesCustomer Experience
Customer Experience Agents
Compare evaluated agents across support tasks, performance metrics, and review signals.
Sort: Evaluation score
Gorgias
4.6534 Reviews
Customer support platform built around ecommerce help desk and automation workflows.
Evaluation scoreA 0–100 summary of the available task evaluation. Use the task results and dimensions to understand what shaped it; this is separate from the G2 review rating.
Hands-on, rubric-based G2 evaluation.
71
G2 tested
Agent evaluationTask-based measures from the named evaluation source. Compare products only when they were tested with the same task set and methodology.
Workflow completionHow often the agent completed the required task outcome—not simply whether it produced a response.
G2 tested
71%
Response accuracyHow factually correct the response was against the scenario context, available knowledge, and supplied policy.
G2 tested
70%
Policy complianceHow consistently the agent followed required rules, confirmations, authorization boundaries, and scope.
G2 tested
100%
G2 reviewsAggregated buyer feedback from G2 reviews. These signals add operational context but are not inputs to the task-based Evaluation score.
Ease of useBuyer-reported ease of using and administering the product. This is review context and does not change the Evaluation score.
G2 Reviews
100%
Quality of supportBuyer-reported satisfaction with vendor support. This is review context and does not change the Evaluation score.
G2 Reviews
92%
Tidio
4.71,754 Reviews
Live chat and AI support automation for small and mid-market customer service teams.
Evaluation scoreA 0–100 summary of the available task evaluation. Use the task results and dimensions to understand what shaped it; this is separate from the G2 review rating.
G2 testing unavailable; provided by Swept.
70
Partner benchmark
Agent evaluationTask-based measures from the named evaluation source. Compare products only when they were tested with the same task set and methodology.
Workflow completionHow often the agent completed the required task outcome—not simply whether it produced a response.
Partner benchmark
70%
Response accuracyHow factually correct the response was against the scenario context, available knowledge, and supplied policy.
Partner benchmark
72%
Policy complianceHow consistently the agent followed required rules, confirmations, authorization boundaries, and scope.
Partner benchmark
92%
G2 reviewsAggregated buyer feedback from G2 reviews. These signals add operational context but are not inputs to the task-based Evaluation score.
Ease of useBuyer-reported ease of using and administering the product. This is review context and does not change the Evaluation score.
G2 Reviews
92%
Quality of supportBuyer-reported satisfaction with vendor support. This is review context and does not change the Evaluation score.
G2 Reviews
94%
Fin
4.586 Reviews
Intercom's AI agent for resolving customer questions across support channels.
Evaluation scoreA 0–100 summary of the available task evaluation. Use the task results and dimensions to understand what shaped it; this is separate from the G2 review rating.
Hands-on, rubric-based G2 evaluation.
66
G2 tested
Agent evaluationTask-based measures from the named evaluation source. Compare products only when they were tested with the same task set and methodology.
Workflow completionHow often the agent completed the required task outcome—not simply whether it produced a response.
G2 tested
66%
Response accuracyHow factually correct the response was against the scenario context, available knowledge, and supplied policy.
G2 tested
55%
Policy complianceHow consistently the agent followed required rules, confirmations, authorization boundaries, and scope.
G2 tested
76%
G2 reviewsAggregated buyer feedback from G2 reviews. These signals add operational context but are not inputs to the task-based Evaluation score.
Ease of useBuyer-reported ease of using and administering the product. This is review context and does not change the Evaluation score.
G2 Reviews
76%
Quality of supportBuyer-reported satisfaction with vendor support. This is review context and does not change the Evaluation score.
G2 Reviews
90%
How scoring works
How evaluations work
What we test
Each CX task defines the customer request, the outcome the agent must achieve, and the policies it must follow. The task set is designed around work buyers expect these agents to handle.
Explore the full methodology
Partner benchmarks are labeled and kept separate from G2-run results.