Wealth-Tech Test-Data Coverage Matrix Worksheet
The fastest way to surface a coverage gap is to put it on a grid. Code paths down the rows, demographic / account / event slices across the columns. A check means a synthetic household exercises that intersection. A blank means a bug is waiting. This worksheet walks the team through filling in the matrix and produces the named archetype slices required to close each gap.
What you walk away with
~18 min · 3 sections · 4 fields- A populated coverage matrix you can paste into a Jira ticket or a steering-committee deck.
- A list of named archetype slices required to close each gap.
- A coverage percentage that's defensible and reproducible across releases.
- An entry point into the Edge-Case Coverage assessment.
Scope
Name the feature whose coverage you're auditing.
One named individual accountable for closing the gaps surfaced here.
Code-path × slice coverage matrix
Mark each cell that has at least one synthetic household exercising the intersection. Blanks are gaps; the next-step block routes each gap to a closing archetype.
| Code path / row | Young / accumulating | Mid-career / peak earning | Pre-retiree / catch-up | Retiree / decumulation | UHNW / complex | Underserved / non-traditional | International / cross-border | Gap notes / archetype to add |
|---|---|---|---|---|---|---|---|---|
Core happy path Median household, common configuration | ||||||||
Tax-edge branches Wash-sale, AMT, NUA, QSBS, multi-state | ||||||||
Retirement-edge branches RMD, Roth conversion, IRMAA, NUA, SS earnings test | ||||||||
KYC / onboarding branches ITIN, dual-citizen, mixed-status, trustee-on-account | ||||||||
Behavioral branches Risk-tolerance change, panic-sell, contribution pause | ||||||||
Compliance branches Reg BI suitability, AML typology, GLBA NPI handling | ||||||||
Error / exception paths Validation failure, unsupported case, escalation |
Coverage roll-up
Live percentages computed from the matrix above.
Auto-counted from your matrix above.
Below 50% indicates structural gaps; above 80% is the threshold for pre-launch defensibility.
Close the gaps
Every blank cell is an archetype you don't have. The Edge-Case Coverage Score assessment ranks them by impact, and the WealthSynth Master Corpus has the archetype slices that close most cross-cutting gaps.
Key takeaways
- Coverage is a property of the joint distribution, not the marginal one. A corpus that covers every code path and every demographic separately can still miss every interesting intersection.
- Error / exception paths are the most under-covered row. Most teams forget them entirely.
- International / cross-border is the most under-covered column. Most teams ship US-only and discover non-resident edge cases in production.
- The matrix is most useful as a versioned artifact. Re-run it per release; the diff is your coverage change-log.
FAQ
Why 7×7?
It's the smallest matrix that captures the structural shape of wealth-tech testing without being unwieldy. Add rows for product-specific code paths, but keep columns roughly stable across features so you can compare matrices across the firm's product portfolio.
Should every cell have at least one archetype?
Most should. Some cells are structurally empty — e.g. 'Young / accumulating' × 'RMD branches' is meaningfully empty (RMDs don't apply). The matrix lets you make that explicit rather than implicit.