The FranchiseData Power Index (FPS)

1.1 Paid Sponsorships & Unverified PR

Many legacy franchise awards and directory lists blend editorial judgment with paid advertorials or self-reported marketing surveys. They rarely audit these claims against statutory Federal Trade Commission (FTC) franchise disclosure documents (FDDs) or audited SEC 10-K filings.

1.2 The Footprint Trap: Gross Size Disguises Failing Stores

Most restaurant industry tables rank brands using one of two blunt metrics:

Because Total System Revenue is simply store count multiplied by average sales per store:

Under traditional lists, a massive legacy chain with 15,000 sub-scale stores doing $500,000 each and closing hundreds of locations a year receives an overwhelming ranking advantage. Meanwhile, an exceptional compounding machine like Chick-fil-A (~3,200 locations generating $8.9M per store) or Raising Cane's (~800 locations generating $5.4M per store) is penalized simply for operating fewer, vastly more profitable stores.

1.3 Ignoring Construction Costs & Payback Speed

A restaurant generating $4.0M in sales sounds impressive—until you discover it cost $9.0M in initial buildout CapEx, requiring over a decade just to break even on invested capital. Conversely, a compact $1.5M revenue box that costs only $450,000 to build pays for itself in under two years. Traditional rankings evaluate top-line revenue while completely ignoring the denominator: the capital required to build the box.

1.4 Concealing Store Closures & Fleet Attrition

Franchise sales marketing routinely highlights new store openings while concealing permanent closures and operator terminations. By ignoring statutory 3-year store census tables (FTC FDD Item 20), traditional lists fail to alert investors when a brand is quietly contracting under the surface.

To balance sheer brand size against individual store profitability, the FranchiseData Power Score evaluates systems across four core financial pillars:

2.6 Calibrated Factor Normalization & Composite Scoring

To ensure stable and mathematically rigorous comparisons across different business models, each raw factor is mapped onto a bounded continuous scale ($[0,100]$):

The Fundamental Base Score (${\text{FPS}}_{\text{base}}$) combines these normalized pillars using calibrated weights, anchored to an institutional baseline:

A ranking engine that moves on every press release is worse than useless: it degrades decision-making. Under the FranchiseData framework, no news event or headline is allowed to change a brand’s score unless it has a direct, measurable financial impact on store cash flows or corporate solvency.

3.1 The Statutory Causation Gate: Why News Gets 0.00 Points

A ranking score should measure business durability, not headline frequency. In traditional media lists, a brand with a large PR agency receives points simply for generating press mentions.

3.2 Verified Transmission Channels

Only three operational transmission channels meet the FranchiseData causation standard:

3.3 The Causation Admissibility Matrix

Institutional benchmarks face a fundamental design challenge: how to reflect real-time developments without introducing erratic noise or subjective point-scoring. The FranchiseData Power Index solves this by maintaining a strict structural separation between audited balance-sheet fundamentals, continuous public equity market discovery, and contextual narrative intelligence.

4.1 Continuous Market Discovery (Public Equity Alpha)

For publicly traded parent enterprises and pure-play restaurant operators, institutional capital markets absorb material operational developments in real time. Rather than relying on arbitrary point adjustments from text articles, the Power Index measures live market alpha (${\alpha }_{\text{live}}$) directly from continuous equity tape feeds relative to the broader restaurant sector beta (${\beta }_{\text{sector}}$):

4.2 Discrete Audited Fundamentals: Why Headlines Never Add Fake Points

Unit economics (AUV, buildout CapEx, net fleet growth, royalty structures) are anchored strictly in statutory regulatory disclosures:

4.3 News as Narrative Context: Explaining the "Why" Without Corrupting the Math

Under our zero-noise standard ($M_{\text{news}}=0.00$), news headlines and regulatory notices never arbitrarily inject or decay points on a synthetic timer. Instead, our automated financial intelligence monitors statutory dockets, SEC 8-K filings, and major operator events to attach real-time narrative catalyst badges to each brand:

4.4 The Final Composite Score

The final live Power Score integrates 100% econometric fundamentals (${\text{FPS}}_{\text{base}}$) with bounded live market discovery (${\alpha }_{\text{live}}$):

5.3 How FranchiseData Compares to Traditional Rankings

How do we know the raw inputs powering the Power Index are accurate? Rather than trusting self-reported marketing surveys, the FranchiseData pipeline enforces a three-tier verification hierarchy: statutory federal filings, mathematical cross-validation (triangulation), and automated pre-build CI invariant assertions.

6.1 Statutory Regulatory Foundations

Every data variable is anchored in legally binding regulatory disclosures subject to federal securities and franchise fraud laws:

6.2 Mathematical Triangulation & Cherry-Picking Gating

We never accept a standalone reported number without cross-validating it against sister variables. The primary mathematical check is Implied AUV vs. Disclosed Item 19 AUV:

Cherry-Picking Detection: If a franchisor discloses an Item 19 AUV of $3.5M, but their $100M system sales across 100 operating units yields an Implied AUV of only $1.0M (a 3.5x ratio), the pipeline automatically detects that the franchisor excluded low-performing stores or sampled only top-quartile units. Any deviation outside the $[0.35\text{x},2.50\text{x}]$ corridor triggers an immediate automated validation halt.

6.3 Automated CI Invariant Assertions (npm run audit:data)

Before any build or production deployment is permitted, our automated test suite evaluates all 538 tracked brands against strict mathematical invariants: