Nikita Boyarkin
← Volta Neobank

Market & Jobs · project 13 of 23

Job segments and behavioral cohorts are not independent (chi² p<0.001): Dormant concentrates in Digital Newcomers 45+ (39.4%) vs Family Budgeters (15.1%).

Situation
The first Market & Jobs project: do job segments (JTBD) match the behavioral cohorts from segmentation
Task
Check whether job-to-be-done segments line up with the behavioral cohorts that segmentation produced
Action
Cross-table JTBD segment × cohort; chi-square for independence and a two-proportion z-test for Dormant; UX-friction contrast
Result
JTBD segment × cohort are NOT independent (p<0.001); Dormant 39.4% among Digital Newcomers 45+ vs 15.1% among Family Budgeters
Stack
Pythonpandas / NumPySciPy / Statsmodelsscikit-learnMatplotlib / Seabornuv + ruff
Source View on GitHub →
On this page
  1. Situation
  2. Task
  3. Actions
  4. Result
  5. Recommendations
  6. Documentation

Volta — JTBD × Cohorts

Situation

The first Market & Jobs project: do job segments (JTBD) match the behavioral cohorts from segmentation.

Task

I opened the market and jobs layer by checking whether segments defined by the job to be done line up with the behavioral cohorts that segmentation produced.

Actions

  • Cross-table JTBD segment × cohort.
  • Chi-square for independence, two-proportion z-test for Dormant.
  • UX-friction contrast (support tickets, KYC duration).

Result

  • Chi-square: JTBD segment × cohort are NOT independent (p<0.001).
  • Dormant 39.4% among Digital Newcomers 45+ vs 15.1% among Family Budgeters (z-test significant).
  • Dormant 45+ show more UX friction → dormancy is driven by UX, not a missing job.

Recommendations

  • The retention lever is assisted onboarding and UX simplification, not accepting churn.
  • Don’t scale one playbook across all job segments.
  • Test the assisted-recovery hypothesis (Project 22).

Documentation

Charts

Source: github.com/NikitaBoyarkin/volta-banking — 22 projects; figures recomputed from the repo's own datasets (data/*.csv) via its analysis scripts. Funnel counts from data/volta_funnel_data.csv (10,000 users); A/B, retention, segmentation, churn, RFM, CLV, attribution, anomalies, spend, support, NPS, JTBD, unit economics, premium, KYC deep-dive, referral, assisted CAC, FX sourcing, premium offers, anchor CAC and dormant win-back follow the published project narrative (README + part pages).

JTBD segment × behavioral cohort

Share of each behavioral cohort within a JTBD segment (%). Job segments and cohorts are not independent: Dormant concentrates in Digital Newcomers 45+.

Young Professionals Digital Newcomers 45+ Travelers Family Budgeters Premium Status Power Growth Casual Dormant Young Professionals · Power: 30.4% 30.4% Young Professionals · Growth: 40.1% 40.1% Young Professionals · Casual: 24.5% 24.5% Young Professionals · Dormant: 5% 5% Digital Newcomers 45+ · Power: 1.8% 1.8% Digital Newcomers 45+ · Growth: 15.6% 15.6% Digital Newcomers 45+ · Casual: 43.3% 43.3% Digital Newcomers 45+ · Dormant: 39.4% 39.4% Travelers · Power: 24.2% 24.2% Travelers · Growth: 40.3% 40.3% Travelers · Casual: 30.6% 30.6% Travelers · Dormant: 5.1% 5.1% Family Budgeters · Power: 7.8% 7.8% Family Budgeters · Growth: 30.5% 30.5% Family Budgeters · Casual: 46.7% 46.7% Family Budgeters · Dormant: 15.1% 15.1% Premium Status · Power: 55.4% 55.4% Premium Status · Growth: 30.3% 30.3% Premium Status · Casual: 11.3% 11.3% Premium Status · Dormant: 3.1% 3.1% 0 100%
Key takeaways
  • Chi-square: JTBD segment × cohort are NOT independent (p<0.001) — job segments ≠ cohorts.
  • Dormant concentrates in Digital Newcomers 45+ (39.4%) vs Family Budgeters (15.1%) — the two-proportion z-test is significant.
  • Implication for the 39.4% Dormant among 45+: the retention lever is assisted onboarding and UX simplification, not accepting churn.

Volta project map

Volta overview →

Core

Extended

Market & Jobs

RAT v2

Causal