Retain_AI/backend/dataForTesting/churn_sample_20k_questionnaire.json at main · Abhayanthk/Retain_AI · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
{
  "_meta": {
    "csv": "backend/dataForTesting/churn_sample_20k.csv",
    "rows": 20000,
    "source": "Kaggle muhammadshahidazeem/customer-churn-dataset (training split, 440k → 20k stratified random, seed=42)",
    "observed_distributions": {
      "overall_churn_rate": 0.569,
      "churn_by_plan_tier": {"Basic": 0.586, "Standard": 0.568, "Premium": 0.552},
      "churn_by_contract_length": {"Monthly": 1.000, "Quarterly": 0.457, "Annual": 0.463},
      "tenure_months": {"min": 1, "median": 32, "max": 60},
      "support_calls_per_customer": {"min": 0, "median": 3, "max": 10},
      "usage_frequency": {"min": 1, "median": 16, "max": 30}
    },
    "auto_detected_columns": {
      "tenure": "Tenure",
      "usage": "Usage Frequency",
      "support": "Support Calls",
      "plan": "Plan Tier",
      "churn": "Churn"
    },
    "what_this_should_exercise": [
      "F1 CoxPH drivers — strong signals: Contract Length (Monthly=100% churn), Support Calls, Tenure, Payment Delay",
      "F2 stat panel — churn_by_plan_tier + churn_by_support_volume + churn_by_usage_decile + tenure buckets all populated",
      "F3 self-consistency — 3 Gemini runs at temps 0.2/0.5/0.7 vote on causes",
      "F4 per-cause RAG — query expansion for each suspected cause",
      "F5 top_segments — Basic plan + Monthly contract should dominate",
      "F9 strategy_skeptic — adversarial review of strategies",
      "F10 RAG-anchored simulation — lift priors from retention frameworks",
      "F11 evidence_dossier — stat → cause → tactic chain",
      "F12 two-pass architect — reasoning_trace populated",
      "F13 HyDE — hypothetical answer for 'first 90 days' segment",
      "F14 competitor_research — fires because churn_destination = HubSpot is in KNOWN_COMPETITORS"
    ]
  },
  "questionnaire": {
    "business_context": "Subscription-based digital service with three plan tiers (Basic / Standard / Premium) and three contract cadences (Monthly / Quarterly / Annual). Customers self-onboard, billing is automated, support is mixed CSM + self-serve. We've grown to ~440k customers historically and currently retain a working sample of ~20k for diagnostic analysis.",
    "industry": "",
    "size": "",
    "business_model": "B2C SaaS",
    "company_stage": "Growth (Series B)",
    "revenue_model": "Monthly subscription",
    "time_range": "",
    "product_lines": [
      "Content library",
      "Weekly digest",
      "Mobile sync",
      "Premium offline mode",
      "Account analytics dashboard"
    ],
    "market_segment": "Working professionals aged 28-48, primarily individual subscribers and small teams, evaluating value monthly and quick to cancel after a single bad billing cycle or support friction event.",
    "budget": "",
    "legal_constraints": [],
    "churn_definition": "Inactivity (30 days)",
    "competitors": ["HubSpot", "Mailchimp", "ConvertKit"],
    "top_channels": ["Paid social", "Organic search", "Referral"],
    "support_model": "Mixed (CSM + self-serve)",
    "goal": "Reduce churn rate",
    "target_churn_rate": null,
    "priority_segment": "Newest customers (first 90 days)",
    "edge_cases": [
      "Monthly contract subscribers are churning at near-100% — we suspect this is partly definitional (the dataset labels them all churn=1 once their contract ends) but want strategies that still treat short-cadence subscribers as a real cohort."
    ],
    "pricing_flexibility": ["Can offer discounts", "Can change plan tiers"],
    "can_ship_changes": "Yes",
    "has_completion_point": "No",
    "churn_destination": "HubSpot",
    "retention_tactics": [
      "Onboarding email sequence",
      "NPS surveys",
      "Win-back campaigns"
    ],
    "timeline": "90-day plan",
    "typical_customer": "Working professionals aged 28-48, primarily individual subscribers and small teams, evaluating value monthly and quick to cancel after a single bad billing cycle or support friction event."
  }
}