Every fulfillment model is a different answer to one trade: cost per order versus control versus your time. There are six answers worth knowing, and the mature version of the decision is per SKU rather than per business, which is why most sellers past their first year run two or three models at once and the real skill becomes keeping the mix visible in one place.
The six models
Self-fulfillment. You store, pick, pack, ship. Cheapest per order at low volume, total control over packaging and inserts, and it costs the one resource you cannot buy back, your hours. Right until the bench time crowds out the work that grows the business, the full math against 3PLs.
Third-party logistics (3PL). Your stock in their warehouse, their team shipping your orders at per-pick rates. Buys back your time and adds geographic reach; costs onboarding effort, storage fees, and a layer between you and the box. The receiving and data disciplines, accurate inbound, live stock feeds, decide whether a 3PL feels like leverage or blindfold.
Amazon FBA. Prime badge, Buy Box advantage, Amazon’s logistics, priced by a fee stack and capacity system that punishes slow stock. Strongest for proven movers on Amazon; the FBA-versus-FBM call is per SKU, not ideology.
Multi-Channel Fulfillment (MCF). The same FBA pool shipping your eBay and store orders, the details that make it work: unbranded packaging, validated tracking, and one shared pool that must stay synced.
Dropshipping. The supplier holds and ships; you hold the listing and the accountability. No inventory risk, thin margins, and availability that is only as true as the supplier feed discipline.
Print-on-demand. Nothing exists until it sells; provider capability is your inventory. Zero stock risk, per-unit costs that cap margin, and the same listing-state control problem as dropshipping.
The decision, run per SKU
Score each SKU on four axes and the model mostly picks itself:
- Velocity: fast movers earn FBA/3PL economics; the long tail rarely does, the ABC lens again.
- Margin headroom: thin margins cannot fund per-pick fees; fat ones buy back your time cheaply.
- Size and weight: oversized items change every fee table they touch, sometimes reversing the answer.
- Promise required: Prime-speed expectations on Amazon push toward FBA; Etsy’s made-to-order patience tolerates self-fulfillment indefinitely.
Run the scoring and the classic hybrid falls out: FBA for Amazon movers, self or 3PL for everything else and the channels FBA cannot serve, dropship or POD for catalog-extension lines you will not stock.
The rule that survives every mix
Fulfillment can split six ways; visibility cannot. Orders from every channel need one queue with routing rules sending each to its fulfillment path, and stock truth needs one synced pool whatever shelf it sits on, because a 3PL count, an FBA count, and a garage count that disagree are three ways to oversell the same unit. The failure mode of hybrid fulfillment is never the models; it is the seams between them.
Common questions
What should a first-year seller use?
Self-fulfillment, almost always: the volume rarely justifies fees, and packing your own orders teaches the product problems (sizing, breakage, returns causes) you need to learn cheap.
When does the switch conversation start?
When bench hours crowd out growth work, when Q4 volume exceeds what your evenings absorb, or when a marketplace’s delivery expectations outrun your postage options.
Can I mix models for the same SKU?
Yes, and profitably: FBA for the Amazon demand, home stock for eBay and your store, with routing deciding per order. The requirement is live sync across both pools.
Which model is cheapest?
Per order at small scale: self-fulfillment. All-in at growing scale: the hybrid the scoring produces, any single model applied to your whole catalog overpays somewhere.
Six models, one truth
Whatever mix fulfills the box, Unifystock keeps one order queue and one synced stock truth across every channel, from $49/month with unlimited orders. See pricing.