FBA, Amazon Warehousing and Distribution, and independent 3PLs are not three competing answers to the same question — they occupy different positions in the supply chain, and most profitable sellers use two of them at once. FBA is downstream fulfillment: it picks, packs, and ships to the customer, and it prices storage aggressively to discourage you from using it as a warehouse. AWD is upstream bulk storage that auto-replenishes FBA, with 2026 base storage around $0.48–$0.57 per cubic foot per month depending on region, plus transportation near $1.26–$1.40 per cubic foot and $1.40 per box inbound and outbound. A 3PL is upstream storage plus its own fulfillment capability, typically $2–$5 per order pick-and-pack and $15–$25 per pallet per month, with the ability to serve any channel. The cost comparison that matters is not storage rate against storage rate — it’s total landed cost per unit sold, including the transfer fees, the prep you now have to buy separately, and the stockout risk each topology introduces.
Why is comparing storage rates the wrong starting point?
Because storage rate is the line where AWD looks best and the line that contributes least to total cost.
<cite index=”24-1″>AWD base storage in 2026 is $0.57 per cubic foot per month in the West region and $0.48 elsewhere — roughly 35–40% cheaper than FBA standard storage and far below FBA’s Q4 peak rate.</cite> On that line alone, AWD wins decisively.
Then the rest of the schedule arrives. <cite index=”24-1″>On top of storage, AWD charges transportation of about $1.40 per cubic foot, or about $1.26 under Amazon-managed distribution, plus $1.40 per box for inbound and outbound processing.</cite> <cite index=”19-1″>Inbound and outbound box processing fees rose by roughly $0.05 per box in 2026, directly incentivizing fewer, fuller cartons and penalizing fragmented cartonization.</cite>
<cite index=”20-1″>A 1,000-unit shipment can incur $250–$350 or more in processing fees before storage even begins.</cite>
And the 2026 increases were substantial. <cite index=”21-1″>West Region AWD storage rose 19% to $0.57/cu ft from $0.48, and the transportation base rate jumped 22% to $1.40/cu ft from $1.15, effective January 15, 2026.</cite> <cite index=”18-1″>Smart Storage rates rose from $0.43 to $0.51/cu ft in the West and from $0.38 to $0.43 elsewhere, with the Amazon Managed rate in the West rising from $0.38 to $0.46.</cite> <cite index=”21-1″>Sellers with California-focused distribution or imports through LA/Long Beach ports absorb the full West-region increase.</cite>
The 2026 change that reprices the entire decision, though, isn’t a fee. <cite index=”20-1″>Amazon discontinued FBA prep and labeling services in the US as of January 1, 2026 — inventory must arrive at AWD already prepped for FBA</cite>, and <cite index=”19-1″>AWD will not act as a downstream prep or correction layer.</cite> <cite index=”20-1″>That’s an additional $0.50–$1.50+ per unit you’re now paying somewhere else.</cite>
If you were using AWD as a low-cost staging layer for factory-direct freight, you now need a prep partner in front of it. That is a 3PL relationship, which means the “one vendor, one dashboard” argument for AWD weakens considerably for anyone whose product needs polybagging, labeling, or bundling.
What does each layer actually cost in 2026?
| Cost element | FBA | AWD | Independent 3PL |
| Storage, off-peak | ~$0.78–$0.87/cu ft/mo (standard) | $0.48/cu ft/mo, $0.57 West | ~$0.35–$0.78/cu ft/mo, or $15–$25/pallet/mo |
| Storage, Q4 | ~$2.25–$2.40/cu ft/mo | Same year-round; off-peak rates extend to Oct 31 with auto-replenishment | Often a peak surcharge, provider-dependent |
| Aged inventory | Surcharge from 181 days, up to ~$6.90–$7.90/cu ft past 365 | Not applicable in the same structure | Long-term surcharges common at 180–240 days, negotiable |
| Inbound handling | Placement fee $0.14–$1.58/unit, $0 with 5+ splits | $1.40/box | $5–$15/pallet ready-to-stow; $25–$45/pallet or $0.35–$1.50/unit for labor-intensive |
| Outbound / transfer | Fulfillment fee from ~$3.20/unit + 3.5% surcharge | $1.40/box + $1.26–$1.40/cu ft transportation | $2–$5/order pick-and-pack, $0.40–$0.50 per additional item |
| Prep & labeling | Discontinued in US as of Jan 1, 2026 | Not offered — must arrive prepped | Standard service, $0.50–$1.50+/unit |
| Carrier cost | Bundled into fulfillment fee | n/a (transfers to FBA) | $5.50–$11.00/order, often with markup |
| Monthly minimum | None | None | ~$517 average |
| Setup | None | None | $250–$1,500 at roughly half of providers |
<cite index=”34-1″>Most ecommerce brands in 2026 pay a 3PL roughly $2 to $3 per B2C order picked and packed (survey average $3.20, B2B about $4.80), $18 to $25 per pallet per month for storage, $5 to $15 per pallet for receiving, and at roughly half of providers a $250 to $1,000 setup fee, before shipping — with monthly minimums averaging about $517.</cite>
Two structural notes on 3PL pricing that change the numbers materially:
<cite index=”40-1″>Storage is often billable two ways, per pallet or per cubic foot, and cubic-foot billing can save 40% for brands with light, high-count SKUs.</cite> Ask which applies before comparing quotes.
<cite index=”34-1″>Location premiums are real — Los Angeles and the NYC/New Jersey metro run 30 to 50% above baseline rates.</cite> A coastal 3PL near your port of entry saves drayage and costs more in storage; the net depends on turn rate.
And a warning that applies to every 3PL model you build: <cite index=”40-1″>invoices run 20 to 50% higher than quoted within six months, and first billing audits surface 7 to 10% in errors.</cite> Model the quote, then budget against the quote plus a variance allowance.
How do you find the break-even between FBA-only and a staged model?
The staged model — hold buffer stock upstream, feed FBA on velocity — costs more per unit in handling and saves on storage and surcharges. The crossover point is a function of turn rate.
Worked example. A 0.20 cu ft SKU, 6,000 units a year, 500 per month average velocity, imported in two container shipments of 3,000 units.
FBA-only: both shipments go straight to FBA. Average inventory held is roughly 1,500 units, or 300 cubic feet.
| Line | Annual cost |
| Storage, Jan–Sep (300 cu ft × ~$0.82 × 9) | ~$2,214 |
| Storage, Oct–Dec (300 cu ft × ~$2.33 × 3) | ~$2,097 |
| Aged inventory exposure on second-container tail | ~$400–$900 |
| Inbound placement, 5+ split via PCP | $0 |
| Total upstream cost | ~$4,700–$5,200 |
Staged via AWD: hold 4,500 units upstream, keep 6 weeks (roughly 700 units) in FBA, auto-replenish.
| Line | Annual cost |
| AWD storage (avg ~450 cu ft × $0.48 × 12) | ~$2,592 |
| AWD inbound processing (~120 boxes × $1.40) | ~$168 |
| AWD outbound processing (~120 boxes × $1.40) | ~$168 |
| AWD transportation (1,200 cu ft × $1.26 managed) | ~$1,512 |
| FBA storage on 140 cu ft rolling position | ~$1,000 |
| External prep, now required (6,000 × $0.75) | ~$4,500 |
| Total upstream cost | ~$9,940 |
On this SKU, the staged model loses — decisively — and the reason is the prep line. Before January 2026, with prep bundled, the same model came in close to or below FBA-only. The prep change is what flipped it.
The staged model wins under three conditions:
- The product needs no prep, or arrives factory-prepped. Remove that $4,500 and the staged model drops to ~$5,440 — roughly at parity, with lower aged-inventory risk.
- Turn is slow or seasonal. <cite index=”25-1″>FBA storage is roughly $2.40/cu ft October through December against AWD’s flat rate year-round. For a SKU sitting in storage through Q4 — which a lot of seasonal inventory does, pre-positioned but not yet selling — AWD saves roughly $1.92/cu ft/month for those three months.</cite>
- You’d otherwise cross day 181. <cite index=”24-1″>AWD’s real value is avoiding FBA aged-inventory surcharges, which start at 181 days and reach roughly $6.90/cu ft past 365 days, along with Q4 peak storage spikes.</cite> <cite index=”25-1″>Add the avoided aged surcharge and AWD is 5–10x cheaper than FBA for buffer stock you’d otherwise hold long-term.</cite>
<cite index=”25-1″>For top-velocity SKUs that turn in 30–60 days inside FBA, AWD adds operational complexity without storage savings — those should ship factory-direct to FBA.</cite>
When does a 3PL beat AWD, and when is it the other way around?
The dividing line is channel mix, not cost.
<cite index=”25-1″>AWD wins on Amazon-internal cost — $0.48 storage against a typical 3PL $0.40–$0.80 — but has friction: case-pack-only inbound, a $0.40–$0.50 per unit transfer-to-FBA fee, and multichannel distribution that costs more than 3PL-native. AWD wins for Amazon-dominant mid-market brands at 60%+ FBA revenue. 3PL wins when you fulfill heavily outside Amazon. The right answer for most multi-channel sellers is to use both.</cite>
<cite index=”25-1″>If most of your demand is Amazon, the AWD-to-FBA transfer flow is your main route, and you’re not paying for multichannel distribution capability you don’t use. If you don’t have a 3PL relationship and don’t want to build one, AWD is the operationally simplest upstream layer — one vendor, one set of dashboards, one inbound flow.</cite>
Three considerations that don’t appear on either rate card:
Rate eligibility isn’t static. <cite index=”19-1″>Smart Storage and managed-rate eligibility is recalculated regularly — a forecasting miss or replenishment dip can degrade rates mid-quarter. Qualification should be treated as an ongoing operational KPI, not a one-time setup.</cite> A cost model built on the discounted rate is a model built on maintaining forecast accuracy.
Inventory loss risk differs. <cite index=”20-1″>AWD’s inventory loss rates run 5–10% according to seller reports.</cite> Reported figures vary widely and should be verified against your own reconciliation, but any shrink rate at that magnitude dwarfs the storage-rate difference between options. A 5% loss on $9 COGS is $0.45 per unit — more than the entire annual FBA-versus-AWD storage delta on a fast-turning SKU.
Every transfer adds a stockout window. A staged model introduces transit and processing time between upstream stock and sellable FBA inventory. Under-provision FBA and you hit the low-inventory-level fee and lose rank; over-provision and you’ve reintroduced the storage cost you staged to avoid.
What about using AWD or a 3PL to serve multiple channels?
Both can, at different price points.
<cite index=”19-1″>AWD is being positioned for multi-channel use by design, with support for custom labels and ASNs enabling flow from AWD to FBA, DTC, and wholesale partners.</cite> But <cite index=”19-1″>parallel increases were flagged across Multi-Channel Fulfillment and Buy with Prime–related costs</cite>, and MCF single-unit fulfillment rose $0.35–$0.41 per unit in January 2026 with a 3.5% surcharge added in May. Serving DTC out of Amazon’s network is now a premium option rather than a cost-saving one, unless your off-Amazon orders are multi-unit — where MCF volume discounts of up to 50% per unit apply.
A 3PL is generally the cheaper multi-channel answer and the more expensive Amazon-only answer. If off-Amazon is a meaningful share of revenue, that’s usually the deciding factor.
Common mistakes in fulfillment cost modeling
Comparing storage rates instead of cost per unit sold. Storage is one line. Transfer fees, processing, prep, and carrier cost are usually larger.
Ignoring the January 2026 prep change. Any AWD model built before 2026 understates cost by $0.50–$1.50 per unit. This is the most common stale assumption in circulation.
Modeling the discounted AWD rate as the baseline. Smart Storage and managed rates are conditional and recalculated. Model the base rate; treat the discount as upside.
Comparing 3PL quotes without normalizing the storage basis. Per-pallet and per-cubic-foot pricing produce wildly different totals for the same inventory. Light, high-count SKUs can be 40% cheaper on cubic-foot billing.
Forgetting the monthly minimum and setup cost. A ~$517 average minimum makes a 3PL uneconomic below a certain order volume regardless of per-order rate.
Omitting shrink. A 3–10% loss rate anywhere in the chain overwhelms every storage-rate optimization on the table.
Treating it as a single decision. Most catalogs want different topologies for different SKUs — factory-direct to FBA for fast movers, staged upstream for seasonal and slow movers. That requires SKU-level cost attribution rather than a channel-level average, which is the reason profit-analytics platforms like sellerboard track fulfillment and storage cost per SKU rather than as a lump operating expense.
Deciding on cost alone. The topology that minimizes cost per unit and the topology that minimizes stockout probability are rarely the same. Rank cost second.
FAQ
Is AWD cheaper than FBA storage? On the storage line, yes — roughly 35–40% cheaper off-peak and far cheaper during Q4. On total cost, only for slow-moving or seasonal inventory, because transportation, box processing, and separately purchased prep offset most of the storage saving on fast movers.
Do I still need a 3PL if I use AWD? Frequently yes, since January 2026. Amazon discontinued US FBA prep and labeling, and AWD requires inventory to arrive already prepped to FBA standards. Products needing polybagging, labeling, or bundling need a prep partner upstream of AWD.
How much does a 3PL cost per order in 2026? Roughly $2–$5 per order for pick-and-pack with $0.40–$0.50 per additional item, plus $15–$25 per pallet per month storage and $5–$15 per pallet receiving. All-in including carrier cost, most benchmarks land around $11–$14 per order.
Does AWD avoid Q4 peak storage rates? AWD storage doesn’t carry FBA’s Q4 multiplier, and sellers using AWD with automatic replenishment to FBA can continue paying off-peak monthly storage rates through October 31, 2026. Inventory that has already transferred into FBA pays FBA peak rates normally.
Can AWD prevent the low-inventory-level fee? Automatic replenishment is designed to keep FBA coverage above the threshold, and reliable replenishment cadence is the main mechanism for avoiding the fee. It’s not a guarantee — the fee is assessed on FBA days-of-supply, so a replenishment lag still triggers it.
Does AWD eliminate inbound placement fees? Routing inbound through AWD is one of the two recognized routes to avoiding placement fees, alongside the Partnered Carrier Program. Confirm current eligibility in Seller Central, since the qualifying conditions have changed more than once.
Should slow-moving SKUs go to FBA at all? Often not. Once storage plus the aged-inventory surcharge exceeds the contribution on a SKU selling one or two units a month, self-fulfillment or a 3PL is the better economic. That threshold is worth calculating per SKU rather than assumed.
How do I know whether my current setup is costing me money? Compare realized fulfillment plus storage plus surcharge cost per unit sold, by SKU, against the same figure modeled under the alternative topology. If you can’t produce that number per SKU, that’s the first gap to close — topology decisions made on aggregate averages are guesses.
Closing
The market talks about FBA, AWD, and 3PLs as though a seller picks one. In practice the profitable configuration is almost always layered: fast-moving SKUs shipped factory-direct into FBA in optimized splits, seasonal and slower inventory staged upstream, and prep handled by whoever does it cheapest now that Amazon no longer does it at all.
What changed in 2026 is that the cheap-storage argument for AWD got substantially weaker — a 19% West-region storage increase, a 22% transportation increase, and prep pushed out of the network entirely. AWD still wins clearly for buffer stock that would otherwise age inside FBA, and it remains the simplest upstream layer for Amazon-dominant brands. But it is no longer the obvious default, and any model built on pre-2026 assumptions is understating it by roughly a dollar a unit.
The number that decides this is total landed cost per unit sold, calculated per SKU, with shrink and stockout risk priced in. Sellers who have that figure make this decision in an afternoon. Sellers who don’t tend to make it once and then live with it for two years.
