Capital Allocation in CPG: Where the Next Dollar Earns the Most

Roughly $54,000 per million. In Cinderhaven Provisions (a synthetic $25M specialty food brand built to model channel economics end to end, not a real company) the next $1M of revenue routed through retail channels produces about $510,000 in contribution. The same $1M routed through distribution produces about $456,000. The company is invented; the cost structures behind the gap (deduction schedules, channel-specific wholesale pricing, compliance penalties) mirror the ones real brands carry. The gap is structural: the two channel types convert a revenue dollar into contribution at different rates.

Most brands never see this number. They allocate capital by revenue rank: the channel generating the most gross revenue gets the most investment. The problem is that revenue rank and contribution-per-revenue-dollar rank are different lists. A brand chasing the first list while ignoring the second is sending incremental dollars to the wrong place.

Revenue rank and contribution rank diverge at the top

Consider the model's channel P&L. Walmart generates $10.8M in gross revenue, the highest of any channel. After eight categories of deductions (promo billbacks, pricing errors, short ships, label fines, spoilage, damaged goods, pallet fines, late delivery), COGS, promotional costs, and dispute overhead, Walmart contributes $5.2M at a 48.4% contribution margin. Whole Foods generates $9.8M in revenue, third place, but contributes $5.5M at a 55.5% margin. First in revenue is third in contribution. Third in revenue is first.

Every board deck built on revenue rank points the business toward Walmart. Every capital allocation model built on contribution margin points it toward Whole Foods. The two are not close. Whole Foods contributes 7 percentage points more per revenue dollar than Walmart does, and the gap compounds with every incremental dollar allocated.

This is not a Walmart problem. It is a structural feature of channels that carry layered cost structures. Walmart's OTIF program charges 3% of COGS per non-compliant shipment against a 98% threshold. Add spoilage, pallet fines, and the dispute overhead required to manage the deduction tail (Walmart collects through compliance rather than slotting), and the full cost-to-serve erodes what looked like the best channel in the company. The OTIF compliance burden is not just an operations problem, it is a capital allocation input.

The 5-point structural gap between retail and distribution

Retail channels earn roughly 51 cents of contribution per revenue dollar. Distributors earn 46 cents. The gap is five points, and it is structural.

Distributors buy at lower wholesale prices than retailers. COGS consumes a larger share of every distributor dollar before contribution is calculated. Distribution looks lower-friction, fewer deduction categories, no slotting fees, no compliance fines, no label penalties. The deduction waterfall for a retailer has eight to ten line items; a distributor waterfall has five. But the friction savings do not offset the wholesale price gap.

A brand evaluating channels by operational complexity reaches the wrong conclusion. Distribution is simpler to manage. It is also structurally less profitable per dollar deployed. The brands that default to distribution growth because it feels easier (fewer chargebacks, fewer portal logins, fewer compliance programs to monitor) are choosing operational comfort over capital efficiency.

This matters most at the inflection points. When a brand wins a new retail authorization and must decide between funding the retail launch or expanding volume through existing distributors, the capital allocation question is not "which is easier?" It is "which returns more per dollar?" On contribution per revenue dollar, at the model's margin structures, the answer is retail, by roughly $54,000 per million. What the contribution comparison does not capture is the capital cost of retail entry: slotting, free fills, the inventory float behind net-60 terms. Launch economics prices that entry ticket; the channel comparison here starts after it is paid.

Volume scales the problem, not the margin

Walmart's marginal contribution per unit declines as volume grows. At 150,000 units annually, the marginal contribution is $11.91 per unit. At 300,000 units, it drops to $10.59. At 600,000 units, $6.01. At approximately 865,000 units, Walmart turns margin-negative in the model, an artifact of how Cinderhaven's promotional commitments and compliance penalties scale, and the exact crossover point will differ brand to brand. The shape is what transfers: deduction rates are not flat in volume, so trade spend and chargebacks can scale faster than revenue.

This is the scale trap. Walmart's buyer asks for more volume. The brand's sales team treats the ask as good news, more units means more revenue. But the deduction rate schedule is not linear. Promotional commitments, compliance penalties, and spoilage all increase as a share of revenue above certain volume thresholds. The brand ships more, invoices more, and keeps less per unit shipped.

The implication for capital allocation is direct. A dollar invested in incremental Walmart volume at current levels returns $11.91 per unit in marginal contribution. The same dollar invested in the same product at Whole Foods, where the contribution margin is 55.5%, returns more, and the curve does not decline the same way, because Whole Foods lacks the layered compliance penalty structure that drives Walmart's marginal cost upward.

Demand forecasting built on revenue targets misses this entirely. A forecast that says "ship 50,000 more units to Walmart" without modeling the marginal contribution at that volume tier is telling the brand to grow revenue while shrinking margin.

The framework is a worked example, not a theory

I built Where the Money Comes From (an interactive capital allocation analysis for a synthetic $25M specialty food brand, Cinderhaven Provisions) to make this math tangible. Five chapters walk a CFO from the revenue illusion (where Walmart looks dominant) through the margin gap, the deduction waterfalls, the scale trap, and finally the capital allocation question itself.

The data covers ten channels: six retailers (Walmart, Kroger, Whole Foods, Sprouts, Costco, and a regional group), three distributors (UNFI, KeHE, DPI Northwest), and DTC. Every deduction category, every COGS allocation, every contribution margin is derived from Cinderhaven's FY2024-2026 channel P&L. The numbers are synthetic. The cost structures are real.

The final chapter puts two scenarios side by side. Route $1M of revenue through retail at the blended 51.0% contribution margin across six retail channels. Route the same $1M through distribution at the blended 45.6% margin across three distributors. Retail yields $510,035 in contribution; distribution yields $456,446. The delta ($53,589 in the model) is the annual cost of allocating by revenue rank instead of contribution rank.

The point is not that distribution is wrong. UNFI and KeHE provide reach that no retail account matches. The choice between DSD and warehouse distribution depends on where the brand sits in its growth arc. But treating distribution growth as the default answer to every investment question (because it is lower-friction, because the deduction tail is shorter, because the sales team already has the relationships) leaves $54,000 per million on the table.

See your own numbers

Every brand's margin structure is different. The 5-point gap between retail and distribution contribution margins is Cinderhaven's number, not a universal constant. A brand with aggressive Syndigo or 1WorldSync data syndication costs layered into its retail channels will see a different spread. A brand whose distributor deduction recovery rate is below average will see distribution look worse than it does here.

The framework is the same regardless. Pull the full cost stack per channel, trade spend, deductions by type, compliance chargebacks, promotional commitments, COGS at the channel-specific wholesale price, and dispute overhead. Normalize to contribution per dollar deployed. Compare.

Lailara builds channel-level capital allocation models for specialty food brands: the same structure behind the interactive analysis, applied to your actual channel P&L. The deliverable is a contribution-per-dollar-deployed ranking across every channel, with the deduction waterfall broken into structural trade costs and recoverable waste. Thirty minutes: tell me which number you are chasing, and I will tell you which one you are actually missing. Book a scoping call.