Methodology
How Viably turns public data into your personal viability score.
Viably doesn't own any of the underlying cost-of-living data. We synthesize 14 public data sources from the current source registry, then turn them into a per-ZIP, per-household viability estimate with visible source vintage, granularity, fallback path, and confidence.
This page is the complete walkthrough. Every formula, every source, every caveat. If a number in your audit looks wrong, this is the page that tells you why.
What's new: Dataset freshness and provenance pass - current official vintages, complete source snapshots, and refreshable metadata.
- ·BLS CPI temporal and regional factors now refresh from the public BLS API and carry the 2026-04 release through record-level source_snapshot_date metadata.
- ·ACS rent, utility allowance, and SMOC paths now use 2020-2024 ACS 5-Year table-based summary files with explicit source vintages instead of stale hardcoded labels.
- ·BEA state RPP ingest now parses all SARPP component line codes, so all-items, goods, rents, and services are aligned to the 2024 official release.
- ·Seeded dataset versions now persist all COL source families, including USDA, EIA, AAA, KFF, MIT, ACS utility allowance, and SMOC.
Try it · 10 seconds
See your viability score before reading the rest
Enter a ZIP and a monthly gross income. We'll run the same calculation explained below and show you the number. No signup.
App calculations
Safe-week, paycheck, and alert math
The app surfaces calculation receipts for the numbers most likely to change a real decision. These receipts use the same backend contracts as the dashboard, forecast, document review, and notification center so a number has one explanation everywhere it appears.
Dashboard safe-week
Safe-to-spend starts with cash on hand, subtracts bills due before payday, adds confirmed paycheck timing, and preserves negative cash instead of hiding overdrafts. The dashboard receipt lists each input, freshness warning, and assumption.
Paycheck forecast
Gross pay comes from shift hours, base rate, premium pay, overtime, and differentials. Net pay subtracts configured pre-tax deductions, estimated taxes, and post-tax deductions. Missing YTD values remain unknown instead of being forced to zero.
Paystub reconciliation
Reconciliation compares extracted paystub actuals against the enriched shift estimate. The core formula isvariance = extracted_paystub_actual - enriched_shift_estimate. Material drift is flagged for review and can calibrate future forecast confidence.
Predictive shortfall alerts
Early warnings fire only when the current week is not already short but projected post-bill cash falls below a bill-load cushion:projected_cash < max(75, min(300, monthly_bills * 0.08)). The alert carries the exact triggering facts into web and future native notification surfaces.
Step 1
The data
Viably queries the current public source registry. Each source has a specific role — housing rents from HUD, regional price differences from BEA, inflation adjustments from BLS, etc. Every source links back to the original publisher so you can verify the numbers yourself.
State-level Regional Price Parities (RPP) from the Bureau of Economic Analysis. Index where US=100.
Metro-area (CBSA) Regional Price Parities from the Bureau of Economic Analysis. Index where US=100. Covers major metropolitan areas; falls back to state RPP otherwise.
County/metro-level Fair Market Rents (40th percentile 2BR). Current vintage: FY2026 revised (effective Oct 1 2025), with FY2024 fallback for any missing counties.
Regional CPI adjustments by Census region. Rescaled from 2026-04 BLS CPI-U release.
ZCTA-level median gross rent anchored to ACS 5-Year estimates (2020-2024). Calibrated using BEA RPP state rents and HUD FMR county data for sub-state variation.
Year-over-year CPI-U indices for temporal adjustment of older source vintages. Current monthly snapshot: 2026-04.
Average residential retail electricity price by state, cents/kWh.
Average retail gasoline price (regular) by state, $/gallon.
County-level monthly cost estimates for typical single-adult living expenses. Derived from MIT's Living Wage Calculator.
State-level median tenant-paid utility + fuel cost, computed as median_gross_rent - median_contract_rent from ACS 5-Year. Used to strip the utility component out of ACS/HUD gross-rent figures so it isn't double-counted against the utilities benchmark line.
Official USDA Food Plans. Per-person monthly food cost at four plan levels (Thrifty/Low-Cost/Moderate/Liberal) by age+sex band. Used as the PRIMARY benchmark for the groceries category. RPP-goods layered on as regional adjuster only.
State-level residential natural gas price multiplier vs US average. Applied to the gas subcomponent (~20% weight) of the utilities category.
State-level median monthly owner housing costs including mortgage, property tax, insurance, condo fees, and owner-paid utilities. Used for the owner-cost pathway (housing_type='owner') parallel to the renter pathway. Utility allowance stripped to produce shelter-only figure comparable to the renter-side contract rent.
State-level average benchmark (second-lowest-cost silver) marketplace plan premium for a 40-year-old. The single most-cited healthcare price benchmark. Used as the primary state adjuster for the healthcare category.
The complete list is available live at /api/col/sources — Viably's own API is publicly readable for this endpoint so auditors can pull the current registry without an account.
Step 2
Housing — the biggest cost category
Housing is ~35% of a typical household's budget. Accuracy here matters most. Viably uses a 3-tier fallback: ZIP-level when available, county-level otherwise, state-median as the worst case. Each tier is named explicitly in your audit.
Start at your ZIP code
We look up HUD Small Area Fair Market Rents (SAFMR), which covers 38,000+ ZCTA-level rent benchmarks across the US. SAFMR is the HUD-published 40th-percentile gross rent for each ZCTA, updated annually (FY2026 is current).
If the ZIP isn't in SAFMR (rural areas mostly), we fall back to HUD county FMR (3,142 counties). If neither tier matches, the state-median applies.
Convert 2-bedroom → 1-bedroom
HUD publishes 0BR–4BR. Our default is the 2BR rate × 82% (the observed national 2BR→1BR ratio). Household size scales this up to the appropriate bedroom tier: singles get 1BR, couples get 1BR, 3-person households get 2BR, etc.
Strip utilities from gross rent
HUD FMR and Census ACS rent are gross-rent sources, meaning tenant-paid utilities are included. Viably subtracts the state tenant utility allowance before the value enters the housing line, because utilities are modeled separately.
Apply CPI-U Shelter inflation
When the upstream source is older than 12 months (e.g. ACS 2022 rent data), we inflate to the current quarter using the BLS CPI-U Shelter index. This adjustment is surfaced explicitly in the audit — you'll see both the anchor value and the CPI-inflated result.
If your result shows "CPI-U Shelter (2025-Q4)" in the housing audit, this is the step that ran.
Scale for household composition
The base rent assumes 1 adult. Additional adults scale 1.6× shared-housing; each child adds a bedroom when household size crosses a threshold (3, 5, 7 people).
Step 3
Non-housing costs
Utilities, food, transportation, healthcare, and "other" use a shared pipeline anchored on BEA's Regional Price Parities.
Formula
category_monthly = national_baseline
× BEA RPP (metro or state)
× BLS Regional CPI factor
× household scaling factor
Utilities
Baseline uses a national monthly utility basket. Metros with higher electricity rates (EIA state series) pay more. Winter-heating states (Northeast, Midwest) see an annualized smoothing.
Food
USDA Thrifty/Low-Cost/Moderate/Liberal meal plans anchor the national baseline; BEA RPP shifts it by metro. Your household size multiplies the per-person figure.
Transportation
Anchored on the BLS CE survey transportation average. AAA state gas prices feed the fuel component; BEA RPP scales insurance and maintenance. Transit-heavy metros (NYC, SF) have a lower vehicle-ownership assumption.
Healthcare
Baseline uses a single-adult employer-coverage healthcare basket. coverage. MIT Living Wage supplies the county-level multiplier. Healthcare premium deductions from your inputs reduce this line.
Step 4
Taxes — gross to net
Viably computes federal + state + FICA in that order, using 2025 IRS brackets. Pre-tax deductions (401k, HSA, health insurance premiums) are subtracted before AGI; post-tax deductions come off the take-home.
Pipeline
gross income
− pre-tax deductions → AGI
− standard_deduction → taxable income
× federal brackets (2025 IRS) → federal tax
+ state rate (per state) → state tax
+ 7.65% FICA (SS wage-base → FICA
capped at the 2025 official wage base)
− post-tax deductions
= monthly take-homeEvery term above is surfaced in your viability audit's "How your take-home was computed" drawer.
Step 5
Data freshness
Public datasets update on irregular schedules — HUD SAFMR is annual, BLS CPI is monthly, AAA gas is daily. Viably runs a refresh scheduler that ingests each source on its own cadence and re-seeds the composite col_records table whenever any upstream source changes.
Every composite release is versioned
Each time col_records is reseeded we publish a new version (e.g. v2026.04.23). Your audit is stamped with the exact version that was live when you ran it — so if the numbers look different next week, you can trace it back.
Every source vintage is tracked
When a source bumps (e.g. HUD FY2026 → FY2027), we log both the before and after vintages. Users who saved an audit get a "your data just got fresher" notification; old audits keep their original numbers for audit trails.
Historical snapshots are queryable
"Which dataset was live on 2026-03-15?" is answerable via /api/col/snapshot-at?date=YYYY-MM-DD. Useful for reconciling old paychecks against the numbers we showed at the time.
Step 6
What we cannot tell you
Viably is a rigorous estimate, not a quote. Our numbers are benchmarks — your specific rent, your specific grocery bill, your specific tax situation may differ. Here's what to keep in mind:
- We use the 40th-percentile SAFMR — about 40% of rentals in your ZIP cost less, 60% cost more. If you're shopping above market, add 15-25%.
- Our tax pipeline assumes a W-2 employee. Self-employed, contractor, and multi-state residents should treat the federal + state lines as a rough lower bound.
- Childcare is costed at the MIT Living Wage county median, which can swing 3× within a single metro. If you're using a specific provider, enter their rate directly.
- Healthcare is for a single adult with employer coverage. Family plans, ACA marketplace plans, and Medicaid scenarios require a personalized quote.
- HUD SAFMR is updated annually each October. For 11 months of the year your housing number reflects the PRIOR year's rent market + CPI-U Shelter inflation.
Step 7
Open where it matters, private where it counts
Open (verifiable)
- • Source registry:
/api/col/sources - • Dataset version:
/api/col/meta/version - • Point-in-time snapshots:
/api/col/snapshot-at - • Version history:
/api/col/version-delta - • This methodology page ← you are here
Private (protected)
- • Your inputs never leave your account
- • Your audits are yours — never shared or sold
- • Public share links (opt-in) scrub all owner identifiers
- • View tracking uses anonymized hashes, never raw IPs
- • Share links expire by default and are revocable anytime
Ready?
Enter your ZIP and your paycheck.
See the numbers in under 10 seconds.
Run your audit Free. No credit card. Takes less than the time you spent reading this page.