Skip to main content

When You Actually Need a Second US Warehouse

Izba
November 18, 2026
10 min read
Share:

A multi-node fulfillment US plan earns its cost when a steady share of your orders sits far from your first warehouse and the savings outrun the added inventory and overhead. Order count alone does not answer it. Delivery time, shipping spend, and how full your first node already is decide the timing.

This is the third item on the Operations & Logistics list in our EU to US Expansion Checklist: plan multi-node fulfillment only once regional volume justifies it. This post gives the thresholds that published sources use for multi-node fulfillment US decisions, shows why they disagree, and explains the mistake our guide warns about, which is opening a second node before the first is full.

When does a growing brand actually need a second US warehouse?

Multi-node fulfillment US networks pay off when three tests pass together. First, a meaningful and steady share of customers waits longer than they will accept. Second, the shipping and conversion gains exceed the cost of a second site. Third, your first warehouse has no cheaper fixes left, such as better carrier rates or more efficient use of its space.

Our EU to US Expansion Guide puts the rule plainly: regional volume, not ambition, should justify the second node. It also lists the failure pattern, which is a second fulfillment node opened before the first one was full. If you have not placed your first node yet, start with our post on choosing your first US fulfillment center location.

What volume and delivery-time thresholds do brands use for multi-node fulfillment US decisions?

Published rules of thumb exist, and they come from 3PLs and logistics vendors. The table collects the ones that include numbers. Treat them as screening tests and not as benchmarks.

Signal Threshold cited Source
Order volume Assess a second node above 1,000 orders per month, with steady growth Stord (opens in new tab), October 2025
Shipping cost share Consider it when shipping reaches almost 10% of total revenue, and keep shipping under 15% of average order value Stord
Shipping cost share A signal when outbound shipping tops 12% of revenue ShopAppy (opens in new tab), 2026
Delivery reach A signal when two-day ground coverage is below 60% with no carrier negotiating room left ShopAppy
Customer geography A signal when over 40% of customers sit in a region the current node cannot reach in two days ShopAppy
Size of the new region If the newly covered region is under roughly 10% to 15% of orders, the math usually says wait ShipAid (opens in new tab), September 2026

ShopAppy adds that one signal is a yellow flag and two is the green light to model a second node. Stord notes that nearly 90% of consumers will wait a maximum of three days, which fits a McKinsey survey (opens in new tab) finding that about 90% will wait two to three days. For delivery time, that makes three days the working line, and our post on US delivery norms for EU brands covers what shoppers expect.

Why do the volume thresholds disagree so much?

Because volume is a weak test by itself, and each source measures something different. Stord says to assess a second node above 1,000 orders a month. ShopAppy says that under 1 million parcels a year, one Midwest node is usually enough. An eFulfillment Service post (opens in new tab) quotes a Red Stag threshold of 50 to 100 orders a day, or $5 million in GMV, below which multiple sites can mean reduced efficiency.

Those numbers span about two orders of magnitude, from roughly 1,000 orders a month to roughly 83,000. Even the coverage estimates vary. ShopAppy puts two-day ground coverage for a single node at 45% to 55% in one article and 25% to 35% in another, and the same publisher's two-node figures also differ between pieces. The lesson is to model your own ZIP codes and not to trust any single number.

So use volume to decide when to start modeling, and use delivery time, cost share, and regional order share to decide whether to act. A brand with 1,500 orders a month spread evenly can sit well inside the window with one node. A brand with the same volume and a third of its customers waiting five days may not.

What does a second node cost beyond the building?

The cost of multi-node fulfillment US networks runs past the lease or the 3PL onboarding fee. ShipAid's model lists setup costs for the site, systems integration, and initial inventory, which it suggests amortizing over 12 to 18 months. It adds a permanent per-order overhead on top, so the cost does not end once the setup is paid for.

Inventory is the largest hidden item. Each site needs its own safety stock, and the eFulfillment post gives the example of a slow item needing one pallet in one warehouse and three pallets across three. ShopAppy, in a separate piece on multi-warehouse fulfillment (opens in new tab), says splitting across two nodes adds 10% to 25% more total units in practice, against about 40% from the square-root rule of thumb. That rule comes from inventory pooling theory, and Stanford research (opens in new tab) argues the effect can differ when demand has heavy tails.

Inbound freight rises too, since stock now travels to two sites. The eFulfillment post models a 75% increase in inbound freight, though its figures are hypothetical and the post comes from a 3PL that favors single-site fulfillment. For an EU brand, that means a second inbound shipment from Europe, which brings its own importer of record and customs filing. Our post on EIN and TIN requirements for EU importers covers who can act as importer of record.

Tax is the last item. Avalara notes (opens in new tab) that storing inventory in another state can create physical presence nexus, so a second node can add a second state to your registrations. Our post on sales tax nexus vs VAT explains how that works.

Why is opening a second node before the first is full a mistake?

Because you pay for duplication before you have used what you already own. "Full" has two meanings here. One is physical, when the first building runs out of space. The other is operational, when the first node has no more headroom in carrier rates, order density, or reach.

Opening early splits your stock into shallower pools. Each pool is more likely to stock out, and a stockout at one site can force an expedited shipment from a farther site or a transfer between warehouses, as the eFulfillment post describes. You also add fixed costs before order density supports them. ShopAppy says zone-skipping only pays off above roughly a few hundred parcels per lane per day, and below that, consolidation and linehaul minimums can outweigh the savings.

ShopAppy's list of what not to do reads like a checklist of early-expansion errors. Do not add a node to serve rare distant customers, since those orders can ship from the existing node at a higher zone. Do not mirror every SKU everywhere, and stock only fast movers in each node. Do not buy buildings before adding order-routing software, or the savings may not show up.

Cheaper fixes come first. Renegotiate carrier rates on the first node, and trim slow SKUs that take up space. Ship the long tail of distant orders from the existing node at a higher zone. Our guide makes the same point about sequencing: move inventory only once the move is justified.

How can you test a second node before you commit?

Pilot one region with your best sellers before you commit to multi-node fulfillment US capacity. Racklify's decision criteria (opens in new tab) suggest mapping order density by ZIP, calculating delivered cost per order by region, modeling the service gain, and piloting one region first. ShipBob (opens in new tab) advises stocking only top sellers at added sites so slow items do not tie up space.

ShipAid gives a rough formula. Upside is the orders from the newly covered region, times a conversion lift, times average order value, plus support savings from fewer "where is my order" tickets. Cost is the setup cost spread over 12 to 18 months plus the permanent per-order overhead across all orders. It suggests a conservative lift of 1 to 3 percentage points, and it says to proceed only if upside clears cost with room to spare.

Results can be large when the geography fits. ShipBob reports a case where a food brand raised two-day shipping coverage from 32% to 65% of its customers and saw a 13% bottom-line saving. That is a vendor-reported case, so use it as an illustration of what is possible and not as a forecast.

How do you decide when to add multi-node fulfillment US capacity?

Work through five steps, and keep the numbers from each.

Step 1: Measure the share of orders outside your delivery window

Pull ZIP-level orders and map them to delivery days from your current node. Count the share that waits more than three days. Note how that share is trending month over month.

Step 2: Check shipping as a share of revenue and order value

Calculate outbound shipping as a percentage of revenue and of average order value. Compare both with the 10%, 12%, and 15% lines in the table above. Two signals together is the point to model a second node.

Step 3: Look for headroom in the first node

Ask whether carrier rates, space use, or SKU trimming could fix the problem more cheaply. If the first building still has room, use it.

Step 4: Model the full cost of the second node

Model multi-node fulfillment US costs in full: setup spread over 12 to 18 months, per-order overhead, extra safety stock, a second inbound leg, and any new state registrations. Apply a conservative conversion lift. If the new region is under roughly 10% to 15% of orders, the math usually says wait.

Step 5: Pilot one region and set a review date

Stock best sellers only, add order routing before adding buildings, and run the pilot long enough to see real delivered costs. Write down the numbers that would justify the next step, and review them on a fixed date.

Frequently asked questions

When does a growing brand actually need a second US warehouse?

When a steady share of orders waits longer than customers accept, the shipping and conversion gains exceed the full cost of a second site, and the first warehouse has no cheaper fixes left. Our guide says regional volume, not ambition, should justify it. A multi-node fulfillment US plan should follow demand, not growth targets.

How many orders per month before a second warehouse makes sense?

No single number holds. Published rules of thumb range from assessing a second node above 1,000 orders a month to staying with one node below roughly 1 million parcels a year. Use volume to start modeling, and use delivery time and cost share to decide.

Does a second warehouse double my inventory?

Not usually. ShopAppy says splitting across two nodes adds 10% to 25% more total units in practice, against about 40% from the square-root rule of thumb. The increase depends on how many SKUs you mirror, so stock only fast movers at the second site.

Does a second warehouse create sales tax obligations in another state?

It can. Avalara notes that storing inventory in another state can create physical presence nexus, as can owning or leasing a warehouse. Ask a US tax adviser which registrations a new node would trigger before you place stock.

What should you do next?

Download the EU to US Expansion Checklist (opens in new tab) and work through the Operations & Logistics list, including the multi-node check. Then read the fulfillment section of the EU to US Expansion Guide to plan your multi-node fulfillment US timing.

This post is general information and does not replace logistics, tax, or customs advice. Thresholds in published sources vary widely, so model your own order data with your 3PL and confirm tax and customs questions with a US tax adviser and a licensed customs broker.

Share:

Related Insights

Ready to Transform Your Operations?

Get personalized insights and a custom growth roadmap from operators who've scaled hundreds of brands.

Free 30-min strategy session • No commitment required