Direct-to-Consumer

How do you compare channels properly?#

1. Map the transaction#

For each channel, write down who sets the displayed price, collects payment, owns inventory, pays fulfilment, handles returns, provides support, and controls customer data. Contract wording and operational reality routinely differ; use the operational reality.

2. Build a channel contribution bridge#

For D2C:

contribution per order = net revenue − product cost − fulfilment − outbound shipping subsidy − payment fees − expected returns cost − variable support − variable acquisition cost

For wholesale:

contribution per unit = wholesale net revenue − product cost − freight and handling − allowances − chargebacks − variable sales cost

Use the same returns, discount, and inventory-loss assumptions on both sides. Both channels are typically recorded net of estimated returns, but wholesale is additionally net of markdowns, allowances, and co-operative advertising — so "net revenue" is not the same construct in the two columns unless you force it to be.

3. Add working capital separately#

Model inventory deposits, production lead times, in-transit goods, processor reserves, refund timing, and wholesale receivables. Contribution margin and cash conversion are different problems and a channel can win one while losing the other.

4. Measure cohorts, not blended averages#

Track first-order contribution, repeat contribution, return rate, time to second order, and CAC by channel and by acquisition cohort. Repeat orders from customers acquired three years ago cannot justify this quarter's paid spend.

5. Price the control you are buying#

Channel control can improve testing speed, pricing latitude, bundling, service, and customer learning. Attach evidence to each claim — faster iteration cycles, higher repeat rate, better attach, lower return rate, higher willingness to pay — rather than asserting the benefit qualitatively.

6. Design the portfolio, not the purity test#

Stores reduce fit uncertainty and absorb returns. Retail partners add reach and credibility. Marketplaces serve search-driven demand. The real decision is usually a channel mix with managed conflict, not D2C versus everything else.

Worked example#

A brand sells a product for $120 on its own website. Per order:

$120 − $32 − $7 − $9 − $4 − $10 − $3 − $38 = $17

The same unit sells wholesale for $58, with freight, allowances, and variable account cost of $6:

wholesale contribution = $58 − $32 − $6 = $20

D2C has more than double the selling price and lower first-order contribution. The case has to be made on repeat. If 35% of D2C customers place a second order with no new CAC and the same non-acquisition variable costs:

second-order contribution per acquired customer = 35% × ($120 − $32 − $7 − $9 − $4 − $10 − $3) = 35% × $55 = $19.25

expected two-order contribution = $17 + $19.25 = $36.25

Read the caveats before you use this number. It is stated before fixed overhead, inventory financing, and any third order. It assumes the repeat order carries zero incremental acquisition cost, which is optimistic if you retarget. It compares contribution per order with contribution per wholesale unit, which is only valid at one unit per order — the channel decision also depends on volume, and wholesale usually moves more units. And the 35% repeat rate is the entire argument: at 20% repeat, expected two-order contribution falls to $17 + $11 = $28, below wholesale on a per-unit basis before any fixed cost.

What are the common mistakes?#

• Calling the price gap "margin." The difference between a $120 retail price and a $58 wholesale price is not profit; fulfilment, returns, payments, service, and acquisition consume most of it.
• Using blended CAC against mature repeat revenue. Match acquisition spend to the cohort it produced, and report payback by cohort.
• Treating first-party data as an asset by itself. Data creates value only when consent, quality, analysis, and action change a decision. An unused customer table is a liability with a storage bill.
• Under-modelling returns. At ~19% of online sales industry-wide, returns deserve their own line, their own forecast, and their own reduction programme.
• Assuming D2C must stay exclusive. Stores, retail partners, insurance relationships, and marketplaces can expand access and improve economics when channel conflict is managed deliberately.

When does D2C break?#

Order economics are too thin for parcel logistics. Low average order value relative to shipping and return cost is fatal, and free-shipping thresholds only move the problem into basket-building.

Repurchase is infrequent or fit uncertainty is high. If every customer must be bought once and buys once, D2C is an acquisition business with a product attached.

Inventory risk overwhelms the margin. Demand is overestimated, lead times prevent reaction, and returns convert reported sales into logistics cost and impaired inventory.

Platform dependency persists despite the direct checkout. Search engines, social networks, app stores, payment processors, and carriers can change prices, policies, or reach unilaterally — see distribution channels.

Direct operations distract from where the brand actually wins. The Honest Company's 2025 decision to stop selling on its own site is the clean example: a brand can conclude that retail partners serve its customers better and that running a storefront is complexity it is paying for twice.