Dropshipping CFO services: margins measured per SKU, not guessed per store

Margin truth per SKU: the waterfall your dashboard hides

A dropshipping store's blended margin is a fiction created by averaging winners against losers, and it hides which is which. We build the waterfall per SKU: selling price, less product cost, less per-unit freight, less duty and brokerage, less payment processing, less a refund and chargeback allowance from your own history, less ad cost per purchase. Only what survives all of that is contribution — and it is common to find active SKUs that contribute nothing once the full stack of costs is applied. Three lines get missed most often:

• Refund lag — refunds land weeks after the sales they reverse, flattering this month at next month's expense.
• Chargeback fees — each dispute costs a processor fee on top of the lost sale, and elevated rates threaten the processing account itself.
• FX spread — selling in USD, paying suppliers in USD or CNY, and banking in CAD means conversion costs hit every leg unless the flows are structured deliberately.

The duty line deserves particular attention now. With US de minimis gone, parcels that once entered free are assessed duties and carrier brokerage charges, and whether you or your customer eats a surprise fee at the door directly hits refunds and repeat rates. Landed cost is no longer optional math; it is the product decision. The compliance side — Canadian GST/HST on your sales, US state sales tax as volumes grow — is covered on our cross-border page for dropshippers.

Testing budgets with kill criteria set in advance

Product testing is the R&D budget of a dropshipping business, and unmanaged, it will eat every dollar the winners make. Two rules fix most of the damage: cap monthly testing spend as a share of last month's contribution, and write the kill criteria before the first ad runs — because after money is spent, every human brain argues for one more day. An example gate framework, with thresholds set per store from your own baselines:

The discipline is the point, not the exact thresholds. A store that kills fast recycles its testing budget several times while a hesitant competitor is still hoping.

Supplier terms are a finance lever, not a procurement chore

Once a SKU proves itself, the cheapest capital available is usually your supplier's. Moving from paying 100% up front to a deposit-and-balance split keeps most of your cash until goods ship; adding a quality holdback or defect credit turns your refund history into negotiating material. Currency of settlement matters too — a Canadian buyer paying a Chinese supplier in USD while selling in USD has a partial natural hedge, and we map those flows before recommending any conversion strategy. Suppliers extend terms to customers with order history and predictable volume, which is one more return on doing the per-SKU math well.

The pivot to US-warehoused inventory — modelled, not guessed

Moving proven SKUs into a US 3PL is the natural next step: bulk freight with duty paid once at formal entry on wholesale value, domestic delivery speeds that lift conversion and cut refunds, and no doorstep surprises for the customer. But it converts a zero-inventory model into a working-capital business, and that trade should be computed, not vibed. We build the comparison per SKU — minimum-order cash outlay, freight and duty, 3PL receiving, storage and pick-and-pack fees against the current direct-from-supplier landed cost — then find the monthly unit volume where warehousing wins. SKUs above the breakeven earn a warehouse slot; everything else keeps shipping direct. Delivery-speed effects on conversion get tested with your own data, not assumed from someone's course. The model also has to absorb a timing reversal: shipping direct, you were paid before the supplier was; holding stock, you pay months before the customer does — and that reversal, not the 3PL fee schedule, is the real price of faster delivery. If the pivot leads you toward FBA, the same modelling carries over — see our Amazon seller CFO services.