Operational Capability Ranker
Methodology

What we measure, how we measure it, and what we don't claim.

The Operational Capability model is a research framework for comparing firms by the structure of what they can execute. Every score in this app is a proxy derived from public filings.

Core formula

Operational Capability is defined as the firm's ability to convert available inputs and relationships into executable work, modulated by the friction of coordinating that work. The multiplicative form is:

Operational Capability =
    Resource Capacity × Relationship Access × Leverage
    ─────────────────────────────────────────────────
                    Coordination Cost

A time-based variant decomposes the same idea into time saved versus time lost:

Firm Operational Capability =
    Resource Time Gain
  + Relationship Time Gain
  + Leverage Time Gain
  − Coordination Time Loss

Variables and proxies

Resource Capacity

Usable inputs the firm can bring to bear.

XBRL tags
CashAndCashEquivalentsAtCarryingValueAssetsNetCashProvidedByUsedInOperatingActivitiesPaymentsToAcquirePropertyPlantAndEquipmentResearchAndDevelopmentExpenseInventoryNetPropertyPlantAndEquipmentNet

Relationship Access

Repeatable access to customers, suppliers, partners, ecosystems, governments, and channels.

XBRL tags
RevenuesDeferredRevenueAccountsReceivable
Text keywords
customersupplierpartnerecosystementerprisegovernmentdeveloperdistributionplatformcontract

Leverage

Asymmetric ability to change options, permissions, costs, prices, rewards, or decisions.

XBRL tags
GrossProfitOperatingIncomeLossNetCashProvidedByUsedInOperatingActivitiesRevenues
Text keywords
pricing powerswitching costsnetwork effectsproprietaryexclusiverecurring revenueplatformecosystemscalebrand

Coordination Cost

Time, money, effort, compliance, and friction required to make cooperation happen.

XBRL tags
SellingGeneralAndAdministrativeExpenseOperatingExpensesRestructuringChargesAccountsPayableCurrentInventoryNet
Text keywords
delaysupply chaindisruptioncomplianceregulatorylitigationrestructuringintegrationdependencyconcentrationcybersecurityrecallmaterial weaknesslabor disputelogistics

Normalization

Every proxy is normalized to a 0–100 score across the current working set, using percentile rank by default (smallest → 0, largest → 100) or z-scoreas an advanced option. Coordination Cost is inverted before entering the composite: higher raw friction lowers the score.

Variable scores combine their normalized sub-components with fixed internal weights (e.g. Resource Capacity = 0.30·Cash + 0.25·OCF + 0.20·Assets + 0.15·R&D + 0.10·Capex). The user-facing weight sliders then scale the four (or eight, in time mode) variables that feed the composite formula.

Confidence score

Confidence reflects the reliability of the underlying data, not the strength of the business. It blends three components:

  • Filing recency (50%) — newer 10-K / 10-Q filings score higher.
  • Text evidence (30%) — number of identified 10-K signals per company.
  • XBRL coverage (20%) — penalty proportional to the count of missing standardized tags.

Confidence is surfaced as High (≥75), Medium (55–74), or Low (<55). Use it as a discount on the headline score, especially for cross-industry comparisons.

Limitations

  • Public filings omit substantial operational detail; many "true" capability inputs are private.
  • XBRL tagging is inconsistent across companies, taxonomies, and fiscal calendars.
  • Keyword-based proxies are coarse — they signal emphasis in narrative, not magnitude of effect.
  • Cross-industry comparisons should be interpreted with caution; the model is most useful within sector cohorts.
  • This is a research proxy, not an audited measure of internal capability, valuation, or investment merit.

Data pipeline (planned backend)

  1. Resolve ticker → CIK via the SEC Submissions API.
  2. Fetch Company Facts XBRL data with a configured User-Agent and exponential backoff.
  3. Use Company Concept for targeted concept refreshes and Frames for cross-company snapshots.
  4. Cache responses server-side; persist companies, filings, XBRL facts, and text signals in Supabase.
  5. Run the scoring engine on demand and store materialized rankings keyed by scoring config + fiscal year.