Dropshipping vs Wholesale: How to Choose Without the Guesswork

Short answer: dropshipping keeps your cash out of inventory and hands fulfillment to someone else, and you pay for that with thin margins and almost no control over what lands on the customer’s doorstep. Wholesale buys you margin, shipping speed and packaging control, and you pay for that by putting money into stock that may not sell. Most sellers who last past year one end up running both — dropship to find out what sells, then buy the winners by the case. The decision that actually matters is not which model to pick, it is knowing the point at which a specific SKU has crossed from one column to the other, and surviving the operational traps on the way across. Every figure below is sourced, and every one of them is worth re-checking before it goes into your cost model.

Dropshipping vs wholesale at a glance

Factor	Dropshipping	Wholesale
Cash per new SKU	Effectively zero — you pay after the customer pays	The full case or MOQ, paid weeks before the first sale
Typical gross margin	Low — the supplier keeps the manufacturing spread and the pick-and-pack labour, and you buy at something close to retail minus a seller discount	Higher — you buy at a unit cost that falls as volume rises
Who ships	Supplier, on their schedule and their cut-off times	You or your 3PL, on yours
Delivery speed control	None. You find out about a delay when the customer does	Full. You pick the carrier and service level
Returns	Yours to accept and process, even though you never touched the item	Yours, but the item comes back to a location you control
Import exposure	Per-parcel duty and entry on every order	One entry per container or pallet
Failure mode	A supplier’s mispick becomes a defect on your account metrics, not theirs	Dead stock and tied-up cash
Best used for	Demand testing, long-tail SKUs, seasonal spikes	Proven repeat sellers with stable pricing

Where each option wins

Dropshipping wins when you do not yet know what sells

The real product of dropshipping is not margin, it is information. Listing 40 variants and finding out which three move is worth doing when the alternative is guessing with a purchase order. It also wins for bulky or fragile goods where storing and re-shipping would cost you more than the supplier’s markup, and for seasonal items you would otherwise be holding in February. The catch is that the information is contaminated by the delivery experience it was gathered under. A SKU that converts while the supplier ships it in twelve days is a different SKU once you hold it and ship in two — so your test tells you what sells badly-shipped, not what the product’s ceiling is. Read the ranking, not the absolute conversion rate.

Wholesale wins the moment a SKU becomes predictable

Once a product sells at a steady rate and you can forecast a month ahead, most of the advantages flip. Your unit cost drops, you control the box the customer opens, you can insert your own paperwork, and a return arrives somewhere you can inspect and restock it. Wholesale also wins on anything where delivery speed is part of the offer — you cannot promise two-day shipping on stock you do not hold. What does not flip is your exposure. The day the pallet lands, a forecasting error stops being an inconvenience and starts being money you cannot spend. And the sample is not the order: a supplier who quietly changes a colourway, a zip pull or a cable length between your sample and your first case can eat the entire margin you just bought, and you will discover it while unpacking, not while negotiating. Approve a production sample from the same run, in writing, before you release payment.

The hidden costs nobody puts in the pitch

Duty on every single parcel. This is the big one, and it is recent. Executive Order 14324, signed 30 July 2025, suspended duty-free de minimis treatment for all countries effective 12:01 a.m. EDT on 29 August 2025 — meaning US-bound shipments valued at $800 or less lost their automatic duty-free pass. CBP then published two interim final rules on 24 June 2026 that suspend the de minimis exemption indefinitely across all modes of transport, including a new informal entry process for mail. Separately, the One Big Beautiful Bill Act enacted in July 2025 terminates the statutory exemption outright on 1 July 2027. If your dropshipping model was built on cheap individual parcels crossing the border duty-free, that model no longer exists as of 2026. Price every landed unit with duty in it, and confirm your specific HTS classification with a licensed customs broker before you commit. Classification is where the money hides — two nearly identical-looking items can sit under different headings at very different rates, and a broker’s consult fee is trivial next to being wrong on a SKU you reorder every six weeks.

The marketplace fee stack. On Amazon’s US store as of 2026, a Professional selling account is $39.99 per month and an Individual account is $0.99 per item sold. Referral fees run from 5% to 45% depending on category, with a $0.30 per-item minimum — though Amazon lists exceptions, including Fine Art and Gift Cards, which carry no minimum. Those percentages are charged on the total sale price, not just the item price, so shipping revenue is fee-bearing too. That last detail quietly punishes the classic low-item-price, high-shipping-charge structure: you pay referral on the shipping you charged, then pay the carrier again out of what is left. Check the current schedule for your exact category in Seller Central rather than assuming the mid-range 15%.

Returns, which are worse than most first-year sellers model. The National Retail Federation’s 2025 Retail Returns Landscape put total US returns at $849.9 billion, or 15.8% of annual sales — and online returns specifically at 19.3%. NRF also estimated 9% of all returns are fraudulent. Roughly one in five online orders coming back is not a tail risk, it is a line item. On a dropshipped order, you refund the customer while your supplier’s return window and restocking terms may not match your storefront policy at all. If your storefront promises 30 days and the supplier’s window closes at 14, every return that arrives in that gap is refunded entirely out of your own pocket, at full retail, on an item you will never see. Line the two windows up before you launch, or price the gap in.

The wholesale side has a paperwork threshold too. Under 19 CFR 143.21(a), shipments not exceeding $2,500 in value can generally clear as informal entries. Above that, you are into formal entry, which means a customs bond and a broker. A first wholesale order of $3,000 therefore costs meaningfully more to clear than one of $2,400 — worth knowing before you round up an opening order. Do not solve this by splitting one order into two shipments to sit under the line; deliberately structuring entries to avoid a threshold is exactly the pattern that draws scrutiny you were trying to avoid. Either keep the order genuinely small or budget for the bond and the broker.

Storefront overhead applies either way. Shopify’s US plans as of 2026 are Basic at $25 per month billed monthly or $19 billed annually, Grow at $65 or $49, Advanced at $399 or $299, and Plus starting at $2,300 per month. That is the floor before apps, and dropshipping automation apps are rarely free — the order-routing and supplier-sync tools you will actually need are usually the paid tier, billed per order or per synced SKU, so the cost grows precisely as you succeed.

The mistake that gets dropshipping sellers suspended

Here is the failure I have seen end more accounts than bad pricing ever did: the supplier’s paperwork ships inside the box.

Amazon’s Drop Shipping Policy requires that you are always the seller of record — your name and business information on all packing slips, invoices, external packaging and any other information included with the product — and that any packing slip, invoice or packaging identifying a third-party drop shipper is removed before the order ships. You must also accept and process returns under your own business name. Ordering from a retail site and having it delivered directly to your buyer, so a competitor’s branded invoice arrives in the parcel, is explicitly prohibited.

How to avoid it: get the blind-shipping terms in writing before your first order, not after. Ask the supplier to confirm in the agreement that they will identify you and only you as seller of record, and that they strip their own documentation. Then verify it yourself — place a test order to your own address and open the box. Do that again after any supplier change, warehouse move, or peak-season surge, because that is exactly when a temp packer reverts to the default slip. The version that hurts is not the supplier ignoring the agreement; it is the supplier honouring it for four months and then hiring twelve people in November. One physical test order is cheaper than one appeal.

Which one fits your situation

You have under a few thousand dollars of working capital and no sales history. Dropship. You are buying data, not margin. Early profit will be small; book it as research spend and judge the quarter on what you learned about ranking, not on the margin line.
You have a SKU with steady weekly reorders and a stable supplier price. Price a wholesale buy. Ask for the MOQ, the unit price at that MOQ, the lead time in days, and the landed cost including duty. If the wholesale landed cost plus your storage and handling still beats the dropship cost, and you can sell the MOQ inside your cash cycle, buy it. If the MOQ represents more than about a quarter’s worth of your current run rate, negotiate it down or wait — a case that takes eight months to clear is a loan you made to your supplier.
Your category is heavy, fragile, or slow-moving. Stay on dropshipping longer than feels comfortable. Storage costs scale with cubic volume, not with revenue, and a low-value bulky item can quietly cost more per month to store than it earns per month in margin.
Your competitive promise is speed or unboxing experience. You need wholesale. Neither is achievable through a supplier you do not control, and no amount of listing copy substitutes for the two days you cannot deliver.
You import into the US. Consolidating into fewer, larger shipments got structurally more attractive in 2026, because per-parcel duty and entry now apply where they previously did not. That tilts the maths toward wholesale for any product you sell in volume — and it means any spreadsheet you built before August 2025 is now wrong in your favour, which is the dangerous direction.

A checklist you can run today

Pull your last 90 days of orders and list every SKU with 20 or more units sold. Those are your wholesale candidates. Everything else stays dropshipped.
For each candidate, email the supplier and ask for four numbers: MOQ, unit price at MOQ, lead time in days, and whether the price is FOB or delivered. If a supplier will not answer the FOB question plainly, treat the quote as incomplete.
Build a landed cost per unit that includes duty, freight, and customs clearance — and check whether the order value crosses the $2,500 informal entry threshold.
Look up your exact Amazon referral fee percentage in Seller Central for each candidate category. Do not use a remembered average, and apply it to the total sale price including shipping.
Place one test order through your own storefront to your own address, open the parcel, and confirm no third-party branding or invoice is inside.
Write down your actual return rate for the last 90 days and compare it to the 19.3% online benchmark NRF reported for 2025. If you are well above it, fix the listing or the supplier before you buy inventory — buying a case of a product with a sizing problem just means owning the problem in bulk.
Check whether your cash can cover the wholesale order and still fund ad spend for the length of the lead time. If not, the answer is not yet.

The short version

Dropshipping and wholesale are not competing business models, they are two stages of the same one. Dropshipping is how you discover demand cheaply; wholesale is how you make money on demand you have already proven. The mistake is picking one permanently — either committing capital to unproven SKUs, or staying on dropship margins long after a product deserved a purchase order. Re-run the checklist above every quarter and let the sales data move each SKU across the line. And verify the numbers yourself before you act on them: Amazon’s fee schedule lives in Seller Central, current import rules are on CBP’s site and in the Federal Register, and the returns data comes from NRF’s annual Retail Returns Landscape. All three moved in the last eighteen months. A cost model carrying a rate that expired in August 2025 will not tell you it is broken — it will just keep returning a margin you are no longer earning, order after order, until inventory arrives and the bank balance disagrees.