# Scale or profit? An operational founder decision framework

**Purpose:** decide whether to accelerate, stage, hold, or shrink a growth program. This is a management framework, not investment advice. Thresholds must be fixed before reviewing results; the numerical example is illustrative, not a universal benchmark.

## The governing rule

Do not ask “growth or profit?” Ask:

> Does the next dollar committed to this cohort, location, product, or capacity block have positive risk-adjusted NPV, and can the company survive long enough to learn whether the thesis is true?

Use the smallest independently testable unit. Blended company averages can hide deteriorating new cohorts behind good legacy customers.

$$
\Delta FCFF_t=\Delta EBIT_t-\Delta CashTax_t+\Delta D\&A_t-\Delta Capex_t-\Delta NWC_t.
$$

Compare expected incremental after-tax cash flows with the specified alternative, including the initial outlay. Capex includes maintenance and growth; working capital is operating working capital. Discount at a cost of capital consistent with the cash-flow risks. Count acquisition costs once and do not double-count execution risk through both a risk-adjusted rate and an extra blanket deduction. Any terminal residual must include remaining reinvestment needs or liquidation recoveries. Do not add unpriced strategic option value. Financing survival is a separate constraint; dilution affects the distribution of equity value, not mechanically operating enterprise value.

Accelerate only when the base case is positive, the downside is survivable, and additional scale improves evidence rather than merely increasing sunk cost.

## Step 1 — Define the economic unit and denominator

Choose one unit before calculating margins:

- subscription cohort by acquisition month and segment;
- order/customer cohort by channel and geography;
- store, route, facility, or city launch;
- machine-hour, model inference, vehicle, or production batch;
- seller/buyer pair or local density cell for a marketplace.

Write a one-sentence rule for every cost: “Would this cash cost change within 12 months if this unit were not acquired or served?” Include second-order service costs, returns/refunds, fraud, payment fees, cloud/inference usage, implementation, local operations, warranty, and partner revenue share. Report both direct contribution and fully loaded incremental contribution.

**Stop sign:** the denominator changes when performance worsens—for example, from paying customers to registered users, or from revenue to annualized run rate.

## Step 2 — Build a cohort ledger

For every cohort or expansion cell, track:

| Dimension | Minimum observable | Decision use |
|---|---|---|
| Demand | price, conversion, usage, repeat rate | Distinguishes paid adoption from subsidized volume. |
| Retention | logo/customer and dollar/revenue retention by month | Determines contribution duration and expansion. |
| Acquisition | fully loaded CAC by channel, including sales labor and incentives | Tests whether growth becomes harder at the margin. |
| Contribution | revenue less causally variable service costs | Tests whether a customer helps cover fixed cost. |
| Payback | cumulative cohort contribution versus acquisition/onboarding investment | Links economics to runway. |
| Capacity | utilization, incremental capex, lead time, maintenance | Exposes lumpy or capital-intensive growth. |
| Working capital | receivable days, inventory days, payable days, deferred revenue | Converts accrual growth into cash needs. |
| Quality | support burden, returns, uptime, safety, churn reasons | Detects hidden diseconomies. |
| Competition | win/loss, price concessions, switching, rival response | Tests whether the apparent moat survives scaling. |

For recurring models, calculate retention on both customers and dollars. Net revenue retention above 100% can coexist with severe logo churn if a few large accounts expand; report the distribution and concentration.

For marketplaces, calculate economics on each side and by local network. National gross volume is not evidence of local liquidity.

For hardware or capacity businesses, include maintenance, replacement, yield loss, warranty, and decommissioning—not only bill of materials.

## Step 3 — Reconcile profit, cash, and economic profit

Prepare a bridge every month or quarter:

1. **Revenue** recognized under the accounting policy.
2. Less direct and variable delivery costs = **unit contribution**.
3. Less operating fixed costs and current-period expenses = **operating profit**.
4. Adjust for noncash items and working capital = **operating cash flow**.
5. Less required capex and capitalized internal-use software; add/subtract other necessary investment = **free cash flow**.
6. Recast selected R&D/CAC as an economic asset only where cohort evidence supports a useful life; charge amortization and failure write-offs.
7. Charge all invested capital at the risk-adjusted cost = **economic profit**.

The cash-flow statement separates operating, investing, and financing cash. Equity and debt proceeds extend runway but are not operating cash generation. [SEC guide](https://www.sec.gov/about/reports-publications/beginners-guide-financial-statements)

### Required adjustments

- **Stock compensation:** keep the GAAP expense in operating economics. Because the indirect cash-flow method adds it back, also report dilution and per-share economics. Calling it “noncash” does not make employee labor free.
- **R&D:** U.S. GAAP generally expenses R&D when incurred. An economic capitalization schedule should use only projects with identifiable future benefit, a conservative useful life, and explicit failure write-offs. [FASB summary](https://fasb.org/page/PageContent?bcpath+=+tff&pageId=%2Freference-library%2Fsuperseded-standards%2Fsummary-of-statement-no-2.html)
- **Working capital:** faster growth can consume cash through receivables and inventory or temporarily generate it through upfront billing and payables. Stress the reversal of favorable timing.
- **Financing:** keep new equity, debt, and venture debt separate from operating cash generation. Customer deposits can be operating working-capital cash flows; model their contractual obligations and possible reversals rather than reclassifying them automatically as financing. Include interest, covenants, liquidation preferences, dilution, and refinancing risk in the survival analysis.
- **Leases:** report fixed contractual payments and duration versus customer commitments. A contribution-positive location can still be economically fragile if a 15-year lease supports cancellable monthly revenue.

## Step 4 — Set four gates before scaling

These gates are proposed management heuristics, not necessary-and-sufficient valuation theorems. Financing dependence or temporarily negative contribution alone does not establish negative expected NPV.

### Gate A: customer truth

Pass only if mature cohorts show that customers continue to pay without escalating subsidy.

Pre-register thresholds suited to the model, for example:

- gross retention at month 12 no lower than the validated base cohort;
- price realization within a fixed range after introductory discounts expire;
- no more than a chosen share of revenue from nonrecurring incentives or related parties;
- concentration stress: loss of the largest customer does not exhaust runway before the next decision point.

**Warning signs:** engagement rises but payment does not; retention is shown only on revenue, not logos; annualized run rate replaces recognized revenue; churn is reset after plan migration.

### Gate B: incremental economics

Pass only if mature-cohort contribution repays CAC, onboarding, and incremental capital inside a defensible customer life and remains positive under downside assumptions.

Required outputs:

- CAC payback distribution, not just mean;
- contribution LTV/CAC with churn, gross margin, discount rate, and horizon shown;
- marginal versus average CAC by channel;
- incremental ROIC and economic profit;
- sensitivity to price, churn, utilization, and variable cost.

Do not use $LTV/CAC$ when LTV is an unconstrained perpetuity. Cap the horizon at the shorter of credible product life, contract durability, and observed cohort history; extrapolated months must be visibly separated.

**Warning signs:** new cohorts have worse payback; “adjusted contribution” excludes recurring support or delivery costs; growth requires permanent negative gross margin; only terminal value makes NPV positive.

### Gate C: scalability and competitive advantage

Pass only if expansion does not recreate the cost base or erode the advantage.

Test:

- fixed cost per unit falls after normalizing for quality;
- capacity utilization reaches the planned range before the next block is committed;
- a new geography reaches target economics within a pre-set learning window;
- network density improves matching, price, or retention locally;
- customer acquisition does not rely on an auction channel whose cost rises with spend;
- competitors cannot copy the offer without a structural disadvantage.

**Warning signs:** headquarters margin improves while local cells never mature; every revenue increment needs proportional sales/support staff; service quality deteriorates; regulators or suppliers capture the scale benefit.

### Gate D: survival and governance

Pass only if the company can reach the next falsifiable milestone under the downside case without assuming a higher valuation or open capital market.

Calculate runway under three cases using actual cash payments:

$$
Runway=\frac{Unrestricted\ Cash-Minimum\ Liquidity\ Buffer}{Monthly\ Net\ Cash\ Burn}.
$$

For seasonal or lumpy models, use a monthly cash schedule rather than the ratio. Include debt maturities, lease commitments, purchase minimums, taxes, and working-capital peaks.

Required governance:

- board-approved scale thesis, metric definitions, budget, and kill criteria;
- independent owner for metric quality and cohort reconciliations;
- staged releases of capital tied to operating milestones;
- conflict disclosure for secondary sales, valuation-based compensation, affiliated transactions, and fund-life pressure;
- a named decision-maker authorized to stop expansion;
- postmortems that compare pre-registered forecasts with outcomes.

**Warning signs:** valuation, fundraising, or gross bookings are executive targets while contribution and cash are not; the board receives only blended metrics; founders or investors can change thresholds after misses; financing proceeds are presented as operational validation.

## Step 5 — Choose one of four actions

| Customer truth | Incremental NPV / survival | Action | Capital posture |
|---|---|---|---|
| Strong | Positive, robust, and financeable | **Accelerate** | Fund validated channels/cells; preserve cohort monitoring. |
| Strong | Positive but highly uncertain or capacity-lumpy | **Stage** | Buy information in tranches; use options and explicit stop dates. |
| Weak/immature | Unknown but downside survivable | **Experiment** | Small reversible tests; do not build fixed cost ahead of evidence. |
| Weak or deteriorating | Negative, fragile, or financing-dependent | **Hold/shrink** | Raise price, cut channels, redesign delivery, or exit the cell. |

Accounting profit is not a separate override. A profitable program should not expand when its incremental expected value is negative; a continuation-versus-exit decision must also account for sunk costs and closure obligations, rather than historical ROIC alone; a loss-making program with high-confidence positive cohort NPV may accelerate if it passes survival and governance gates.

## Step 6 — Treat strategic options honestly

Network effects, data, standards, and platform options can justify faster scale, but they need observable milestones.

Write the option claim in falsifiable form:

- “If density in city X reaches Y, median match time will fall below Z and 90-day retention will rise by A.”
- “If installed base reaches Y, attach revenue per unit will cover platform fixed cost by quarter Z.”
- “If model usage doubles, unit inference cost will fall by A% without quality loss.”

Value only the option the current investment actually buys. Discount for technical failure, competitor preemption, regulation, and future financing. A large TAM is not option value unless the firm can access it with a defensible advantage.

## Illustrative worked example — not a benchmark

Assume a B2B subscription cohort of 100 customers:

- Year-one recognized revenue per customer: $12,000.
- Cash CAC and onboarding: $5,000 per customer at inception.
- First-year fully loaded contribution margin: 60%.
- Annual gross revenue retention: 82%; no expansion revenue.
- Required return: 20%; horizon: four years; no terminal value.
- Additional shared-platform spend for the cohort: $100,000 at inception.

Then cohort contribution before discounting is:

| Year | Revenue | Contribution at 60% |
|---:|---:|---:|
| 1 | $1,200,000 | $720,000 |
| 2 | $984,000 | $590,400 |
| 3 | $806,880 | $484,128 |
| 4 | $661,642 | $396,985 |

Initial outlay is $600,000 ($500,000 CAC/onboarding plus $100,000 platform). Discounted contribution is approximately $1.48 million, so illustrative NPV is roughly $0.88 million before working capital, taxes, failure, and competitive response.

Now stress gross retention to 60%, contribution margin to 45%, and CAC/onboarding to $8,000 per customer. Initial outlay becomes $900,000 and discounted contribution falls sharply. Under those specified downside assumptions, discounted contribution is $843,750 and NPV is −$56,250. The founder should not cite the original blended LTV/CAC; this simplified result supports reconsidering the commitment, subject to the omitted taxes, working capital, and risk assumptions.

The point is procedural: lock the variables, show the downside, and let new cohorts update the decision.

## A one-page board decision memo

Every scale request should fit this template:

1. **Decision unit:** cohort/city/product/capacity block.
2. **Counterfactual:** what happens if we wait, shrink, or do nothing?
3. **Customer evidence:** price, retention, repeat use, cohort maturity.
4. **Economic bridge:** contribution → operating profit → operating cash flow → free cash flow → economic profit.
5. **Capital required:** CAC, capex, leases, working capital, and fixed platform costs.
6. **Base/downside NPV:** assumptions, discount rate, and probability weights.
7. **Strategic option:** observable mechanism and expiry date.
8. **Competition:** expected response and evidence for durability.
9. **Runway:** monthly cash low point under base/downside; next financing not assumed.
10. **Milestones:** dates, independent data owner, accelerate/hold/stop thresholds.
11. **Governance/conflicts:** secondary liquidity, valuation incentives, control rights.
12. **Decision:** accelerate, stage, experiment, hold, shrink, or exit.

## Warning-sign dashboard

Escalate to the board when any two occur, or immediately for a solvency/safety issue:

- revenue growth accelerates while fully loaded contribution per unit declines;
- new-cohort payback worsens for two consecutive cohorts;
- retention is deteriorating but masked by price increases or customer concentration;
- variable cost is being reclassified as “growth investment”;
- capex per incremental revenue rises without a verified utilization benefit;
- working-capital outflow grows faster than revenue for reasons other than a planned inventory build;
- stock compensation rises faster than gross profit and dilution is omitted from targets;
- the next round, IPO, or acquisition is necessary to avoid breaching minimum cash;
- secondary sales reward insiders before unit economics are validated;
- the company enters multiple new markets before one new market reaches maturity;
- management changes the cohort window, denominator, or profitability definition after a miss;
- valuation milestones substitute for customer or cash milestones.

## What this framework cannot decide

No dashboard resolves radical technical uncertainty, strategic interaction, regulation, or the private value founders place on control. Judgment remains necessary. The discipline is to expose that judgment as assumptions, cap downside through staged commitments, and prevent financing availability or accounting presentation from being mistaken for evidence of enterprise value creation.