Table of contents
- TL;DR
- The Failure Cycle
- When Two Locations Win
- When It Backfires
- Prerequisites
- 8 Guardrails
- Decision Matrix
- FAQ
Have you ever heard the phrase, “You’re tripping over dollars to pick up pennies?” In a lot of cases, that’s exactly what two-location fulfillment can look like — especially for small parcel brands shipping lightweight products.
AI depiction of a man in a warehouse tripping over dollars to pick up pennies.
⚡ TL;DR (For Operators)
Two locations only improve outcomes when each warehouse ships meaningful volume and demand is predictable enough to allocate inventory confidently.
If you split too early, you will see more stockouts, more transfers, and worse promise accuracy.
The “Two Locations Made Everything Worse” Cycle
When a team says “our second warehouse didn’t help,” they are usually describing symptoms rather than root causes.
| Symptom | Customer Impact | Root Cause |
|---|---|---|
| More backorders on best sellers | Longer waits, more cancellations | Fast movers were split; each node runs out sooner |
| Constant transfers | Delays, extra labor, added freight cost | No allocation rules or replenishment triggers |
| Higher WISMO volume | Support load increases | Checkout promises don’t match node-level reality |
| Inventory “In Stock” but won’t ship | Refunds and trust loss | Location-level accuracy and rules are not tight enough. |
| Shipping cost doesn’t drop | Margin pressure | Volume per node is too low to reduce zones consistently. |
The uncomfortable truth: Any network will look “bad” if it runs without allocation discipline.
Mindset Shift: Two Locations ≠ “More Capacity”
Many brands approach a second location with a simple (and flawed) logic: “If we open another node, everything gets faster.”
A better framing: “Two locations is a network design choice. It only pays off if zone reduction outweighs added inventory risk.”
When Two Locations Are a
WIN
- Meaningful volume by region
- Stable SKU velocity
- Clear split candidates
- Reliable replenishment
When Two Locations Are a
BACKFIRE
- Volume per node is too low
- Fast movers are fragmented too early
- Transfers are not controlled
- Checkout logic isn't location-aware
When Two Locations Are a Win
Two-node fulfillment tends to work when most of the following are true:
- Meaningful volume by region: A large share of your orders consistently comes from a second region.
- Stable SKU velocity: Stable movers are easier to allocate, replenish, and keep in stock.
- Clear split candidates: High-velocity SKUs with low damage risk and predictable replenishment.
- Reliable replenishment: You can handle two receiving workflows without inbound becoming a bottleneck.
When Two Locations Backfire
It’s usually the wrong move when:
- Volume per node is too low: You take on overhead without getting meaningful zone reduction.
- Fast movers are fragmented too early: Splitting a small set of "hero" SKUs causes frequent stockouts.
- Transfers are not controlled: If stock transfers aren't a scheduled process, they become "emergencies."
- Checkout logic isn't location-aware: If the site promises 2-day shipping but the item is 2,000 miles away, you lose.
Prerequisites Before You Split
It’s usually the wrong move when:
- Volume per node is too low: You take on overhead without getting meaningful zone reduction.
- Fast movers are fragmented too early: Splitting a small set of "hero" SKUs causes frequent stockouts.
- Transfers are not controlled: If stock transfers aren't a scheduled process, they become "emergencies."
- Checkout logic isn't location-aware: If the site promises 2-day shipping but the item is 2,000 miles away, you lose.
The Mountainy Method: 8 Practical Guardrails
1. Regional Inventory Allocation Rules
Set allocation targets by region using recent order distribution. If the West region represents 40% of demand for SKU A, allocate 40% of inventory to the West node—only if the SKU is split-eligible.
2. SKU Eligibility Rules
Not every SKU should be duplicated. A-SKUs (Fast) split only with enough depth. B-SKUs split selectively. C-SKUs (Slow) keep single-node to avoid stranded inventory.
3. Safety Stock Per Node
One warehouse needs one buffer. Two warehouses require more total buffer. Do not split a SKU unless each node can hold a minimum number of "days of cover" after safety stock is accounted for.
4. Replenishment Triggers Per Node
Define Min (The "Oh no" trigger) and Max (The cap to prevent over-allocation) for every node.
5. Planned Transfer Policy
Define when transfers happen, who owns them, and how they ship. If it's ad hoc, it's a cost center.
6. Promise-Aware Routing
Routing should respect what the customer saw at checkout. If a node is low, do not route there just because it’s geographically closer if it can't meet the delivery promise.
7. Exception Monitoring
Track stockout rate on top SKUs, split-ship rate, and WISMO per 1,000 orders weekly by node.
8. Phased Rollout
Start with a short list of split-eligible SKUs in one specific region. Expand only after multiple replenishment cycles show stability.
Decision Matrix: “Should We Split This SKU?”
| Checklist Item | Split Across Both Locations If... | Keep Single-Node If... |
|---|---|---|
| Velocity | Consistent weekly movement | Promo-driven spikes or irregular sales |
| Inventory Depth | Each node has deep days of cover | Splitting makes both nodes "thin" |
| Regional Demand | Meaningful demand in both regions | Demand is concentrated in one spot |
| Replenishment | Inbound cadence is reliable | Inbound is late or inconsistent |
The uncomfortable truth: any network will look “bad” if it is run without allocation discipline.
What to Tell Your Team
If you are hearing “we need a second warehouse,” reset the frame:
- “Two locations is a complexity milestone, not just a capacity one.”
- “We split inventory to reduce zones without increasing stock risk.”
- “No rules means more transfers, more stockouts, and broken promises.”
FAQ
Is two-location fulfillment worth it for Shopify brands?
Yes, if regional demand is consistent and SKU velocity supports node-level safety stock.
What is the biggest mistake brands make?
Underestimating how much extra safety stock is needed when you fragment inventory.
Do we need to split every SKU?
Almost never. Only split SKUs with clear regional benefits and stable replenishment.
