The Rule
Every physical product founder knows their margin and their growth rate. Neither answers the question that actually matters. Is this business compounding, or slowly dying while looking busy?
ROIC + Organic Growth ≥ 25%. 155 companies scored. 50 backtested. 80% accuracy for physical retail. Three questions to find your number.
Metrics Fail
The margin illusion
Every physical product founder knows their gross margin and their revenue growth. Their accountant reports it. Their broker talks about it. Their retailer buyer negotiates on it. Those are the numbers the ecosystem gives them.
Neither metric answers the question that determines whether the business compounds or slowly dies. What is this business earning on the capital it has tied up?
A company can show 45% gross margins and 35% revenue growth while destroying value, because those numbers ignore the $800K in inventory, receivables, slotting fees, and compliance infrastructure sitting underneath them.
Four failure paths from the same measurement error
1. The pricing trap
Price set on margin, not on capital return. A 50% gross margin sounds healthy until you realize the capital deployed to earn it produces a 12% ROIC. The founder is pricing against the wrong benchmark.
2. Trade spend burn
72% of promotions lose money when measured on ROIC instead of sales lift. The promo "works" by moving units. It fails by destroying return on the capital funding it.
3. Channel expansion error
All channels look similar on margin. They are 3x different on ROIC. Target and Walmart might both show 40-45% gross margin on the term sheet. The capital required to service each channel, and the effective retailer take after all hidden costs, produces wildly different returns.
4. DTC scaling illusion
Margin stays flat while ROIC collapses. Customer acquisition costs climb every year. The denominator grows. The numerator stalls. Invisible until the cash runs out.
The problem is not that founders use the wrong metrics. Nobody in the founder's ecosystem speaks capital efficiency. The accountant doesn't. The broker doesn't. The retailer buyer doesn't. The consultant doesn't. The equation exists. Buffett uses it. Costco is built on it. It just never reaches the founder in the room.
The math isn't hidden. The equation is just bigger than anyone told you.
Data Shows
Seven investors, six decades, one metric
Across seven independent investor frameworks, 181 company financial analyses, and hundreds of founder and operator case studies, one metric appears with extraordinary consistency. Return on Invested Capital.
| Source | Material | What They Measure |
|---|---|---|
| Buffett | 60 shareholder letters | Owner earnings. "A wonderful company at a fair price." |
| Bezos | 27 shareholder letters | FCF per share. "We optimize for free cash flow, nothing else." |
| Sleep | 24 Nomad letters | Robustness Ratio. "Earn less, but for a much longer time." |
| Leonard | 19 CSI letters | ROIC + Organic Growth. "The ceiling on long-term shareholder returns." |
| Marks | 45 Oaktree memos | Consistency of returns. "27th-47th percentile every year = 4th percentile over 14 years." |
| Goldratt | 4 books | Throughput per constraint. "Improving a non-bottleneck by 50% improves the system by 0%." |
| Mauboussin | Published research | Two paths only. "Low cost or differentiated. No middle." |
Why ROIC, not margin
Margin measures profit per dollar of revenue. ROIC measures profit per dollar of capital deployed. A 3% margin business with almost no capital tied up (Costco) can produce 25% ROIC. A 20% margin business with heavy capital requirements can produce 10% ROIC. Both numbers are accurate. Only one tells you whether the business is efficient.
Two paths to competitive advantage
Mauboussin's framework says there are only two paths. Low margin with high capital velocity (Costco, Trader Joe's, IKEA). High margin with constrained throughput (Hermes, Rolex, AutoZone). Operating margin only captures path two. ROIC captures both.
| Company | Op. Margin | ROIC | Why They Diverge |
|---|---|---|---|
| Costco | 3.4% | ~25% | Negative working capital. Vendors finance inventory. Membership income on zero capital. |
| AutoZone | ~20% | ~40% | Negative working capital by 10 days. 86% share count reduction compounds the math. |
| Trader Joe's | ~6-7% | ~30%+ | 10,000 sqft stores. 60x inventory turns. Zero debt. Cash on delivery. |
| Hermes | Very high | Very high | 70% gross margins. Deliberately constrained production. Both paths produce high ROIC. |
Costco scores 13 on the Rule of 40 (10% growth + 3% margin). The most durable retailer in history looks like a failing business. Costco scores 35 on The Rule (25% ROIC + 10% growth). The metric changes. The verdict changes.
ROIC here means cash profit divided by all capital tied up in the business. Not GAAP margin. Not EBITDA. Actual cash returned on actual capital deployed. Inventory, receivables, slotting, compliance, equipment, working capital. Everything that's in the business and can't be used for anything else. This is a screening lens, not a scientific law. It is the best available threshold for separating physical businesses that compound from those that don't.
Two rules, two domains
The Rule of 40 screens SaaS companies. The Rule screens physical businesses. They use different inputs for the same structural reason. Capital intensity.
SaaS capital intensity is near-zero and uniform. Margin IS the differentiator. Works because the denominator (invested capital) is roughly the same for every SaaS company.
Physical capital intensity varies enormously. Margin alone is meaningless without knowing how much capital generates it. ROIC captures what margin misses: capital efficiency.
Three layers
The Rule is not a single test. It is three layers. Each is necessary. Each answers a different question.
The Velocity Gate
Binary. Pass or fail. Does the product move fast enough to hold its position? Below the minimum velocity threshold for a given retailer, you lose money regardless of unit economics. 95% of new consumer products fail here. You have 60-90 days to prove it.
The Return Test
ROIC + Organic Growth ≥ 25%, with ROIC independently ≥ 15%. The core screen. It captures both paths to competitive advantage. Low margins with high capital velocity (Costco, Trader Joe's) and high margins with constrained throughput (Hermes, Rolex). Mark Leonard's formula, independently validated across Buffett, Bezos, Sleep, Marks, Goldratt, and Mauboussin.
The Constraint Signal
High returns can be temporary. The businesses that last deliberately earn less than they could. Costco caps markup at 14%. Hermes restricts production. Trader Joe's opens 1-2 stores per year. The self-imposed constraint converts operational discipline into customer trust. Trust generates duration. Duration generates compounding.
Why each layer is necessary
Without Layer 1: You run the ROIC math on a product that's about to be delisted. The numbers are irrelevant if the product is dead on the shelf.
Without Layer 2: You have a surviving product with no idea whether it's compounding or treading water. You're Walmart, not Costco. You're Petco, not AutoZone.
Without Layer 3: You have high returns with no durability signal. You're HelloFresh during COVID. You're Peloton in the lockdown. High ROIC driven by favorable conditions, not structural advantage.
Before you calculate anything
The formula only works if you accept the premises underneath it. These six beliefs separate the founders who compound from the ones who look busy. Each one comes with a question you can answer tonight with your own numbers.
Your margin is not what your accountant reports. It's what you take home after everything.
Pull your last 12 months by channel. After co-op, chargebacks, trade spend, freight, and compliance. What's the real number?
Revenue growth and value creation are not the same thing. Growth can destroy a business.
Separate your organic growth from channel launches and paid acquisition. How much growth came from what you already had?
Every channel has a different return on the capital you put into it. Most founders never compare.
Take your total capital deployed into each channel. Divide your actual cash profit by that number. The answers will not be close to each other.
Getting the PO is not the milestone. Surviving the first 90 days on shelf is.
95% of new consumer products fail the velocity window. Do you know your retailer's minimum threshold? Are you above it?
The best businesses deliberately earn less than they could. That restraint is what makes the returns last.
Are you saying yes to every opportunity, or choosing the ones where your capital earns the most?
Nobody in your professional ecosystem is asking the question that actually matters. You have to ask it yourself.
Your accountant, your broker, your buyer, your consultant. Which one of them has ever asked what your capital is earning?
Calculate It
Three questions
The Rule reaches a founder through three questions. Most physical product founders have never been asked any of them.
Not revenue. Not sales. Total capital deployed. Inventory on hand, receivables outstanding, slotting fees committed, compliance infrastructure, equipment, working capital buffer. The pile of money that's IN the business and can't be used for anything else.
Founders say "$200K in inventory." The real number, after walking through receivables, slotting, compliance, and working capital, is usually $500K-$800K. The gap is 2-3x. This is the denominator of ROIC. Getting it wrong means every margin calculation is built on the wrong base.
Not gross margin. Not EBITDA. Actual cash that came back after everything. Channel by channel. After the real retailer take (not the headline margin), after trade spend, after chargebacks, after freight, after returns.
There is a 15-point gap between headline margin and actual contribution margin at major retailers. Target's effective take is ~49%, not the 40% on the term sheet. That "45% gross margin" the founder quotes produces a 25% real contribution. Their $540K in gross profit is actually $180-250K in cash profit.
Not total growth. Organic growth. Same stores, same channels, same products, growing because demand is getting stronger. A new Costco launch is a capital deployment, not organic growth. More ad spend is a capital deployment, not organic growth.
Founders say "we grew 35%." After separating new channel launches and increased ad spend, organic growth is usually 5-10%. The rest is new capital being put to work. The Rule needs the organic number because growth funded by capital deployment is already captured in the ROIC calculation.
Then you do the math. ROIC = Cash profit (Q2) / Capital tied up (Q1). Add Organic Growth (Q3). Above 25% with ROIC above 15%, the business is likely compounding. Below that, the founder needs to understand why before signing the next PO.
Methodology
The denominator is all invested capital: inventory, receivables, slotting, compliance, equipment, and working capital. Not just inventory. Not just "cost of goods." The numerator is actual cash profit after all channel-specific costs, not GAAP gross margin. This is Buffett's "owner earnings," simplified. The organic growth figure excludes new channel launches and increased paid acquisition. These are capital deployments, not organic growth.
Examples
What it looks like in practice
Capital deployed: $400K (inventory $150K, slotting $80K, packaging $50K, trade spend $40K, compliance $30K, working capital $50K). Year-one wholesale revenue: $800K. After COGS and Walmart's actual take, real profit: ~$224K.
Organic growth: 10%. Sum: 66%.
Above threshold. This channel is working.
Capital deployed: $500K (Target's compliance and slotting run higher). Year-one revenue: $600K. Target's effective take: ~49%. Real profit: ~$132K.
Organic growth: 10%. Sum: 36%.
Passing, but half the Walmart return. Same product, same brand. The channel choice alone is the difference.
Capital deployed: $300K (inventory $60K, CAC $200K, fulfillment $40K). Year-one revenue: $500K at 65% gross. After $200K marketing and $40K fulfillment: ~$85K profit.
Year 3: ROIC 18% + Growth 8% = 26%. Barely hanging on.
CAC rises every year. No consumer brand has exceeded $300M through DTC alone.
Same sauce brand. In Walmart at 56% ROIC. Target calls. Amazon pushes hard. The founder says no to both. Reinvests the Walmart capital deeper. More SKUs, more stores, better execution in the existing relationship.
That's the constraint signal. Earning less revenue than they could. Every dollar of capital returns more than twice the threshold.
Run the three questions on the current DTC business AND on the Costco channel separately. Compare ROIC by channel. Does Costco earn a better return than what you're already doing? If DTC is producing 28% ROIC and Costco projects at 45%, the capital is better deployed in Costco. If Costco projects at 15% after all costs, stay DTC and optimize what you have.
The Rule doesn't say "go to retail." It says go where the capital earns the most.
Data first, pattern second
Not a theory. Not a framework invented and then illustrated. Data first. Pattern found. Formula extracted. Tested against outcomes.
How it was built
Seven independent investors (Buffett, Bezos, Sleep, Leonard, Marks, Goldratt, Mauboussin) were analyzed across their complete published works. All seven converged on return on invested capital as the metric that predicts business durability. Mark Leonard stated the formula explicitly in 19 years of Constellation Software shareholder letters. The others confirmed it implicitly from different directions.
155 companies were scored against the Rule from detailed financial analyses. 44% pass rate, meaning the threshold discriminates. A test where everyone passes, or everyone fails, tells you nothing.
How it was tested
50 companies were scored using only the financial data available in 2021-2022. Outcomes were checked 3-4 years later. The scoring was done blind, without knowledge of what actually happened to each company.
80% accuracy for physical retail businesses. 72% across all 50 companies tested, which included SaaS, financial infrastructure, and industrials. The gap exists because the Rule wasn't built for digital businesses. It's a physical-business screening tool tested against a broad dataset to see where it breaks.
Backtest results
| Result Type | Count | Description |
|---|---|---|
| Correct Pass | 25 | Rule said durable. Company thrived. |
| Correct Fail | 11 | Rule said fragile. Company stagnated or died. |
| False Positive | 8 | Rule said durable. Company struggled. |
| False Negative | 6 | Rule said fragile. Company thrived. |
Where it breaks
The 8 false positives all had the same problem. Growth was carrying the score, and that growth was temporary. Not fraudulent. Temporary. HelloFresh had COVID pull-forward. Twilio and ZoomInfo were growing 30-40% by acquiring companies with near-free debt. Calm rode a mental health surge during lockdowns. Floor & Decor and Old Dominion hit cyclical peaks. When growth contributes more than 60% of the combined score, the Rule gets unreliable.
The ROIC floor (minimum 15%) was added to catch this. If ROIC on its own doesn't clear 15%, the business is riding growth momentum, not capital efficiency. The floor would have caught 3 of the 8 false positives. With it applied retroactively, overall accuracy improves from 72% to approximately 79%.
The 6 false negatives had two patterns. First, structural transformations. GE was a broken conglomerate that scored terribly, then broke into three companies and GE Aerospace took off. Netflix was borderline, then pivoted to ad-supported and changed the economics. The Rule scored the business as it was. Reality changed the business into something different. Second, business models outside the Rule's scope. Berkshire Hathaway runs on insurance float. DoorDash and Uber were pre-profit marketplaces building network effects. Current ROIC doesn't capture models where the value is in the network, not the operations.
When to trust it most. ROIC is the dominant component of the score. The business is a physical product company with measurable unit economics. Growth is funded by operations, not external capital or temporary demand.
When to be skeptical. Growth is the dominant component. The growth might be cyclical, subsidy-driven, or trend-based. Or the business model is structurally different from physical retail.
Not just big brands
The core research draws on Costco, AutoZone, Hermes. Large public companies with decades of data. Fair question. Does this work for a brand doing $10M trying to get into Costco in 2026?
We tested against modern brands in today's retail environment. The Rule caught every failure.
Overhead went from 44% to 68.5% of revenue. 34 of 42 stores fell below the velocity threshold. Closing all US stores by early 2026. The Rule saw it. ROIC went negative. Stores were expensive customer acquisition pretending to be a growth channel.
Customer acquisition cost hit $300+ per person. Went to Target and Walmart. Acquired at a fraction of its IPO value. Every dollar of growth made the problem worse because the capital required to acquire each customer exceeded what that customer returned.
Hardware margins collapsed from 40% to under 20%. The demand was cyclical, not structural. The 15% ROIC floor would have flagged this immediately. Growth was temporary demand, not durable economics.
Constant discounting compressed margins until there was nothing left. The channel choice (Amazon-first with aggressive pricing) destroyed the unit economics that made the product viable.
Leases of $300K per year in markets where the brand had low awareness produced an effective customer acquisition cost of $600 per person. Digital ads in the same markets cost $50-100. The remaining 7 stores became profitable. The Rule saw both the problem and the correction. ROIC was below threshold during expansion and above threshold after the stores closed.
Five modern brands. Five failures. All caught.
On the other side, brands that would pass:
Two years of R&D before selling a single bag. $150K in lab testing. Deliberately constrained to specialty retail only. No mass grocery. Fibonacci growth pattern for 8 consecutive years. Passes all three layers.
Went cash-negative for 18 months to invest in R&D when they could have coasted on growth. Brought science in-house. Rejected funding offers worth hundreds of millions. "1% growth in HiSmile is better than 3,000% growth in something else." Constraint signal present on all four proxies.
Developed the hero product over 2 years before showing it to his own board. Personally guaranteed 50,000 bars to the first retailer. Sold out in 2 hours. Now producing 120M+ bars per year in the UK alone, across 80 countries.
Liquid Death ($333M in 4 years across 133,000 retail doors), OLIPOP (~$400M through Costco, Target, and Walmart), Poppi (acquired by PepsiCo for $2B), and Feastables ($250M with 8% margin on near-zero customer acquisition cost) are also directionally consistent. Full verification isn't possible because they're private. But the trajectory fits.
The pattern is the same whether the brand is worth $200B or $2M. Deploy capital into a channel. Earn a return on that capital. If the return plus organic growth clears 25% and ROIC independently clears 15%, the business compounds. If it doesn't, the business is on a treadmill no matter how fast revenue is growing.
This isn't a rule for Costco-sized companies evaluating their empire. It's for the brand sitting across from the Costco buyer, deciding whether to say yes.
The Rule is not a validated scientific law. It is a working screening tool. The 25% threshold originates from software company data (Leonard's CSI) and has not been independently calibrated for physical retail. ROIC calculations vary by methodology. No standard exists across analysts. Prospective validation (predictions locked today, checked in 3 years) has not yet been conducted. The dataset skews toward notable companies. The backtest is retrospective, not prospective.
Where people get it wrong
The sloppy denominator
Inconsistent ROIC calculations produce different verdicts on the same company. If one analyst includes leases and another doesn't, the numbers diverge by 10+ points. The methodology must be locked before the calculation is trusted. Include all capital: inventory, receivables, slotting, compliance, equipment, and working capital.
Confusing total growth with organic growth
Launching in a new retailer is a capital deployment, not organic growth. Increasing ad spend is a capital deployment. "We grew 35%" is usually 5-10% organic after stripping out new channels and new paid acquisition. The Rule needs the organic number because the capital deployed for growth is already captured in the ROIC denominator.
The growth trap
A 40%-growth company with 5% ROIC scores 45% but is not capital-efficient. That's why the floor exists. ROIC must independently clear 15%. A company passing on growth momentum alone is one demand shift away from failure.
The constraint signal is qualitative
Layer 3 cannot be reduced to a single number. Use the four proxies: (1) gross margin below industry while ROIC above, (2) growth below capacity, (3) zero or low debt, (4) employee retention 2x+ industry. At least 2 of 4 should be present. Honest assessment matters. Most founders will claim they're constrained. Look at their actual behavior.
Cyclical exposure
Quality businesses in cyclical industries can look like failures over short windows. John Deere's ROIC drops during agricultural downturns. Measured across a full cycle, it clears the threshold easily. Use 3-5 year averages for cyclical businesses, not point-in-time snapshots.
When The Rule doesn't apply
Pre-revenue companies (no capital to measure returns on). Pure service businesses (capital intensity is labor, not inventory). Regulated utilities (returns are set by regulators, not markets). Holding companies and conglomerates (ROIC is an average across unlike assets). The Rule is for physical businesses with measurable capital deployment and market-set returns.
Tools Connect
The system
The Rule is the diagnostic. The tools are how you fix what it finds.
Shows the gap between perceived and actual margin. When a founder says "45% gross margin," the Mirror shows them the 15-point gap to real contribution. Fixes the numerator.
Shows actual cost structure by channel. Retailer take, trade spend, chargebacks, freight, compliance. The full denominator. When the founder sees what's really deployed, the ROIC math changes.
Addresses the price-setting that determines the numerator. Ten plays for when margin is a mirage. Pricing against capital return, not against cost-plus or competitor benchmarks.
Simulates the velocity gate. Layer 1 of The Rule. 140 scenarios across channels, categories, and price points. Binary. Does the product move fast enough to survive?
The three questions as a structured form. Locks down the methodology. What counts as invested capital, what counts as cash profit, how to separate organic growth. Input your numbers, get your score.
Reference
What was studied
Investor letters and shareholder memos: Buffett (60 annual letters), Bezos (27 letters), Nick Sleep (24 Nomad Partnership letters), Mark Leonard (19 Constellation Software letters), Howard Marks (45 Oaktree memos), Benjamin Graham (111 essays), Eliyahu Goldratt (4 books on constraint theory).
Company deep dives (181): Full financial analyses of public and private companies from Business Breakdowns, covering retail, industrial, consumer, and services. 155 companies scored against The Rule.
Founder and operator case studies (hundreds): Long-form interviews and analyses across physical product, retail, and consumer businesses. Founders, operators, and investors across multiple industries and eras.
Retail industry benchmarks: Pricing structures, margin waterfalls, velocity thresholds, compliance costs, promotional economics, and inventory math across major US retailers.
Published works on economics and strategy: 39 deep analyses spanning Adam Smith, Tocqueville, Madison, Machiavelli, Roger Martin, Ben Thompson, and others. Historical and modern frameworks for institutional durability.
The seven investors and what each measures
Buffett: Owner earnings. Cash generated after maintenance capex. The number GAAP doesn't show you. Applied via "wonderful company at fair price" filter.
Bezos: Free cash flow per share. "Percentage margins are not one of the things we are seeking to optimize." Capital deployed against free cash flow generated.
Sleep: Robustness Ratio. Customer savings divided by shareholder profit. Costco: 5:1. Higher ratio = wider moat, lower current profit, longer duration.
Leonard: ROIC + Organic Revenue Growth. Stated the formula explicitly. "The ceiling on long-term shareholder returns." Applied across 800+ acquisitions at CSI.
Marks: Consistency over magnitude. "27th-47th percentile every year = 4th percentile over 14 years." Returns above cost of capital, sustained across cycles.
Goldratt: Throughput per constraint dollar. "Improving a non-bottleneck by 50% improves the system by 0%." The constraint is the system.
Mauboussin: Two paths only. Low cost (high capital velocity) or differentiation (high margins). No sustainable middle. ROIC captures both; margin captures only one.
Company scoreboard
| Company | ROIC | Growth | Sum | Verdict |
|---|---|---|---|---|
| AutoZone | ~40% | ~5% | 45% | Pass |
| Costco | ~25% | ~10% | 35% | Pass |
| Hermes | ~35%+ | ~10% | 45%+ | Pass |
| Sherwin-Williams | ~25%+ | ~7% | 32%+ | Pass |
| Chipotle | ~35% | ~10% | 45% | Pass |
| Floor & Decor | ~35% | ~15% | 50% | Pass |
| Winmark | ~95%+ | ~10% | 105%+ | Pass (outlier) |
| Walmart | ~15% | ~5% | 20% | Below threshold |
| Petco | ~12% | ~5% | 17% | Below threshold |
| Sears | Negative | Negative | <10% | Fail |
| Bed Bath & Beyond | Declining | Negative | <10% | Fail |
| Peloton | Collapsed | Negative | Volatile | Fail |
Backtest detail
2021 cohort (25 companies): Scored using only financial data available at the time. Outcomes checked 3-4 years later. Physical retail companies (AutoZone, Costco, Floor & Decor, Chipotle, Sherwin-Williams) scored with 80% accuracy. Digital companies (HelloFresh, Twilio, ZoomInfo, Calm) scored worse due to growth inflation from COVID and cheap capital.
2022 cohort (25 companies): Higher accuracy overall due to post-COVID normalization. False positives concentrated in growth-dominant scores where growth exceeded 60% of the combined number.
Where it was wrong: Every false positive shared the same trait. Growth was the dominant component, and that growth was temporary. The ROIC floor (≥15%) catches most of these retroactively. With the floor applied, overall accuracy improves from 72% to ~79%.
Physical retail: 80% accuracy. The Rule discriminates well for its target domain. It was not designed for, and does not reliably screen, digital-first businesses.
The constraint signal and how to detect it
At least 2 of 4 should be present for a durable constraint signal:
1. Gross margin below industry average while ROIC above industry average. The business is sharing economics with customers. (Costco, Trader Joe's, IKEA.)
2. Revenue growth below what unit economics would support. The business is choosing not to grow faster. (Trader Joe's opens 1-2/year. Hermes trains artisans over years. Chipotle is owned-and-operated only.)
3. Zero or near-zero debt. The business funds growth from operations. No external capital provider can force short-term optimization. (Trader Joe's zero since 1976. IKEA 500 kronor. Mars zero. Costco minimal.)
4. Employee retention 2x+ industry average. The business invests in people at rates competitors won't match. (Costco 7% turnover vs. retail 33%. Trader Joe's 5-6%. Hermes 6%.)
More coming soon
Channel-specific ROIC benchmarks. Industry-adjusted thresholds. Prospective validation results.
New tools and research as we build them. No pitches. Just the work.
We'll email when something ships. That's it.
Got it. We'll be in touch.