Channel P&L: Finally Seeing Which Marketplace Actually Makes You Money

Ask a multichannel brand which channel is most profitable and you will get an answer delivered with confidence and constructed from vibes. Revenue dashboards exist everywhere; profit per channel per SKU exists almost nowhere, because the costs live in five systems that do not talk. Building the channel P&L is unglamorous data plumbing with one redeeming quality: it changes almost every decision it touches.

Where the real numbers hide

Fees live in settlement reports, not fee schedules. Amazon alone nets referral fees, FBA fees by size tier, storage (with seasonal multipliers), returns processing, ad spend and a bestiary of adjustments out of your deposit. The published rate card is a theory; the settlement file is the fact — and reconciling them routinely surfaces both errors worth claiming and fee drift worth knowing. Parse every settlement, per SKU, per marketplace. This is the foundation and the reason most brands never start.

Fulfilment differs per channel per unit. FBA for this size tier vs your 3PL pick-pack vs WFS — plus inbound freight allocation. Model each honestly; the differences decide the FBA-vs-merchant-fulfilled question per SKU, not per religion.

Ads must land on the SKU. Channel ad spend allocated to advertised SKUs (and their halo, if you want sophistication) turns TACoS from a channel abstraction into a line on each product's P&L. This is where "our bestseller" gets reclassified as "our most subsidised".

Returns hit the channel that caused them. Rates differ by channel (marketplace fashion returns run structurally higher), and processing costs differ by path. Return-adjusted or it is fiction.

The statement that results

Per channel, per SKU, monthly: revenue → minus refunds → minus channel fees (settlement truth) → minus fulfilment → minus allocated ads → minus returns cost → contribution. Rolled up per channel with fixed channel costs (subscriptions, dedicated headcount, integration amortisation) beneath. Suddenly answerable, with numbers: which channel deserves the next SKU launch; which SKUs to pull from which channels; whether Zalando's volume survives its returns rate; what Amazon's real take rate is (spoiler: brands guess 15%, discover 28–38% after ads and fulfilment).

What changes when it exists

In every engagement where we have built this: SKU-channel pruning that raised total contribution while lowering revenue (a trade every CFO accepts), ad budgets reallocated from vanity SKUs to quiet earners, fulfilment methods flipped per SKU, and at least one "strategic" channel demoted to seasonal. The dashboard pays for itself in the first pruning round; after that it is the steering wheel.

It sits naturally on the right-sized stack: settlement connectors in, dbt models applying the cost waterfall, one screen out. Three to five weeks of build for most brands — and the end of profitability by anecdote.