Churn vs Acquisition Analysis - Freshdesk & Slack InvestigationCustomer & Account Analytics
How each model performed on this task. Every chart states its N, and the reward carries a 95% confidence interval once there is more than one trial to measure.
Core metrics
Check Pass %
Mean share of the task's checks passing, averaged over trials. N = 10 trials per model.
98%
Claude Opus 4.8
91%
GPT-5.5
88%
Gemini 3.1 Pro
38%
Gemini 3.5 Flash
Extended metrics
Median Time
Wall-clock minutes, median across trials. N = 10 trials per model.
4m
Gemini 3.5 Flash
3m
GPT-5.5
3m
Claude Opus 4.8
2m
Gemini 3.1 Pro
Tool Calls
Tool invocations, median across trials — counted from the trajectories. N = 10 trials per model.
32
GPT-5.5
30
Gemini 3.5 Flash
26
Claude Opus 4.8
18
Gemini 3.1 Pro
Verifier × model pass rate
How often each verifier passed, per model. Green is passing, red is failing — a red band across a row is a verifier nothing clears.
| Verifier | GPT-5.5n=10 | Claude Opus 4.8n=10 | Gemini 3.1 Pron=10 | Gemini 3.5 Flashn=10 |
|---|---|---|---|---|
| checks (22) | ||||
| avoided_slack_post_messages | 100% | 100% | 100% | 100% |
| avoided_ticket_status_or_priority_changes | 100% | 100% | 100% | 100% |
| cfo_acquisition_case_in_answer | 100% | 100% | 0% | 0% |
| conversations_1464_customer_reason | 100% | 100% | 90% | 0% |
| conversations_1677_feature_complaint | 100% | 100% | 90% | 0% |
| conversations_1884_closure_request | 100% | 100% | 90% | 0% |
| note_on_ticket_1464 | 70% | 90% | 100% | 0% |
| note_on_ticket_1884 | 70% | 90% | 100% | 0% |
| notes_only_on_1464_and_1884 | 100% | 100% | 100% | 100% |
| reported_closure_ticket_volume_in_answer | 50% | 100% | 60% | 0% |
| retrieved_5star_squad_history | 100% | 100% | 80% | 70% |
| retrieved_growth_squad_ad_budget | 100% | 100% | 100% | 100% |
| retrieved_ticket_1464 | 100% | 100% | 100% | 0% |
| retrieved_ticket_1677 | 100% | 100% | 100% | 0% |
| retrieved_ticket_1884 | 100% | 100% | 100% | 0% |
| retrieved_user_info_person_106452 | 100% | 100% | 100% | 70% |
| retrieved_user_info_person_106486 | 100% | 100% | 100% | 100% |
| searched_closure_tickets | 60% | 100% | 100% | 0% |
| summarized_churn_ticket_reasons_in_answer | 100% | 100% | 90% | 0% |
| summarized_growth_ad_decisions_in_answer | 70% | 70% | 80% | 70% |
| summarized_person_106486_in_answer | 90% | 100% | 60% | 40% |
| summarized_ratings_install_impact_in_answer | 100% | 100% | 100% | 80% |
Failure modes on this task
32 tagged verifier failures, by the stage they were assigned to. Higher is worse.
| Failure stage | GPT-5.5 | Claude Opus 4.8 | Gemini 3.1 Pro | Gemini 3.5 Flash |
|---|---|---|---|---|
| Answer1 mode | 8 | 4 | 10 | 10 |