Skip to content

issue-012.md

Latest commit

 

History

History
101 lines (62 loc) · 7.62 KB

issue-012.md

File metadata and controls

101 lines (62 loc) · 7.62 KB
issue_type Enhancement
linear_workflow Option A — parent issue + child themes (T0–T2 slices in `/create-plan`); extends Phase 4 (issue-011) shipped capabilities — does not replace merchant/UPI/compare/chat foundations
source_plan .cursor/plans/money_mirror_10x_issue_34a61725.plan.md (Money Mirror 10x / Gen Z clarity loop)
prior_cycles experiments/ideas/issue-011.md (Phase 4 P4-A–P4-H shipped), experiments/plans/plan-011.md
linear_root VIJ-52 — https://linear.app/vijaypmworkspace/issue/VIJ-52/issue-012-gen-z-clarity-loop-emotional-ux-frequency-perf-slas
linear_project https://linear.app/vijaypmworkspace/project/issue-012-gen-z-clarity-loop-emotional-ux-frequency-perf-slas-1bbe50903d79
app apps/money-mirror

Issue Title

MoneyMirror — Gen Z clarity loop: emotional UX, frequency-first insights, and performance-to-insight SLAs

Type: Enhancement

Problem (Current State → Desired Outcome)

Current state: After Phase 4 (issue-011), MoneyMirror delivers statement-native ground truth, merchant / UPI visibility, bad-pattern advisories with tap-through, month-over-month compare, facts-grounded coaching and optional chat, weekly recap email, and proactive opt-in surfaces. For the core ICP — Gen Z / young Millennials in India who are UPI- and quick-commerce-native and may be between jobs, dependent on family support, or using informal liquidity — the product can still feel like accurate accounting without a complete emotional safety + habit loop: small debits stay mentally invisible (Swiggy, Zepto, entertainment micro-orders), transactions and dense lists feel slow to resolve into a story, and AI value risks sounding generic unless tightly coupled to their frequency and merchant clusters. Performance (time-to-Mirror-card, time-to-first-advisory) is not yet owned as explicit product SLAs.

Desired outcome: Users experience a shame-reduced, story-first money review: frequency × small-ticket signals and merchant clusters (“quick commerce,” “food delivery ecosystem”) are default-visible alongside totals; guided reflection (short questionnaire or 3-step review) turns data into one commitment for the next statement; notifications stay high-signal (recap + optional threshold) with specific copy tied to facts; dashboard load matches Gen Z speed expectations via progressive disclosure, skeleton-first patterns, and lazy heavy lists. All generative copy remains facts-grounded (Layer A + bounded txn context); no new invented amounts.

User

Primary: Gen Z and young Millennials in India (including ₹20K–₹80K/month band and adjacent) who are digitally fluent but avoidant or unclear about cash flow — especially those in transitions (no salary for months, gig irregularity, or reliance on parents / informal loans) and who lose track of micro-expenses across UPI and cards.

Secondary: Product/owner — Linear parent + themed children after /explore → /create-plan; dependency awareness on issue-011 shipped surfaces (merchant aliases, compare-months, bad-pattern advisories, proactive stubs).

Why This Problem Matters

Outcome before features: The North Star proxy (repeat statement upload ≤60d, issue-009/010/011 continuity) only compounds if users return for meaning, not only accuracy. Trust (PDF truth) is necessary; emotional safety + clarity on “where it all went” drives habit. Without frequency-native and cluster views, competitors that only show category pie charts still win on simplicity of story — Money Mirror must own the story at merchant + behavior level without moralizing.

Opportunity

Position MoneyMirror as the India Gen Z app that combines bank-statement evidence, merchant-native rollups, and a weekly clarity loop tuned for UPI / quick-commerce reality — differentiated from SMS aggregators and investment-first apps. This issue packages UX, narrative, and perf contracts that turn Phase 4 engineering into a 10× felt experience for the ICP.

Hypothesis

If we ship shame-safe copy, frequency-first and clustered-merchant surfaces, guided money review (questionnaire-length flows grounded in Layer A facts), and explicit performance budgets for dashboard/transactions — while keeping weekly/high-signal proactive rules — then engagement depth (Insights, Transactions, advisories), repeat upload, and willingness to pay (paywall experiments) will increase versus totals-only and generic AI narratives.

Theme map (Linear-ready — parent + future children; finalize in /create-plan)

ID Theme Notes
T0 Emotional design + perf UX Voice/empty states; skeleton-first; time-to-Mirror / time-to-first-advisory targets; progressive txn load
T1 Frequency + cluster insights Orders/week, micro-UPI story UX, default “quick commerce / food” clusters on top of merchant_key + aliases
T2 Guided loop + proactive rules 3-step review + optional saved commitment; notification copy rules; tie-in to recap / opt-in channels

Depends on: issue-011 P4-A / P4-E / P4-F (merchant/UPI, bad patterns, compare) as baselines — extend, do not fork.

Non-goals

  • Real-time bank link aggregation or SMS scraping as MVP of this issue.
  • Investment product recommendations or credit underwriting.
  • Shaming, streak-based punishment, or dark-pattern nudges.
  • Replacing deterministic advisories with unconstrained LLM numeric claims.

Success metrics (draft for /metric-plan)

  • North Star proxy (unchanged): repeat statement upload within 60d of first success (cohort).
  • Supporting: guided_review_completed (or equivalent), time-to-first-advisory / dashboard ready timers, merchant cluster or rename engagement, notification opt-in vs churn, existing badpattern_advisory*, merchant_rollup_clicked, recap opens.

Risks / Open Questions (for /explore)

  • Tone: Extend docs/COACHING-TONE.md for dependence / family support without being patronizing — user testing copy.
  • Cluster taxonomy: Who defines default “quick commerce” lists (curated merchant_key sets vs ML) — maintenance cost.
  • Performance: Measuring app-specific timers vs Core Web Vitals only — what’s feasible on Vercel + Next.js 16.
  • Privacy: Guided review answers — storage minimization and opt-in for retention.
  • Dependency: Completing /learning for issue-011 and /linear-close on VIJ-43 vs starting exploration in parallel — PM choice.

Personal testing protocol (repo)

  1. Upload multi-month statements; confirm Mirror, compare, merchant rollups, bad-pattern CTAs still coherent after UX changes.
  2. Throttle simulation: heavy dashboard loads respect rate_limit UX (issue-011 P4-H).
  3. PostHog: single-source events for any new funnel (per product-lessons: instrument during build, not after).

Next step

/explore — validate assumptions (competitor patterns, India UPI micro-spend, shame-safe UX), and record in experiments/exploration/exploration-012.md.