When Should Your Brand Add a Second Warehouse?
Test whether a second warehouse improves real orders after stock duplication, transfers, split shipments and setup work enter the decision.
The Short Answer
Add a warehouse to solve a demonstrated customer or capacity problem, not simply because sales are growing. Map products and common baskets alongside destinations. Compare the current operation, an improved single-site arrangement and a specific two-site plan using the same orders.
Include receiving, storage, transfers, coordination, setup and additional inventory needs beside outbound transport. Test how routing behaves when one location lacks an item, and examine quiet periods and replenishment delays. Pilot a suitable product and regional group before expanding. Judge complete delivery, stock availability, exceptions and actual cost; the number of warehouses is not itself a measure of progress.
Key takeaways
- Define the service or capacity problem a new site must solve.
- Plan stock around complete baskets as well as regional demand.
- Compare full operating scenarios and the assumptions that could reverse the result.
- Test routing and a limited assortment before distributing more inventory.
A map makes a second warehouse look simple: place stock nearer more customers, reduce distance and deliver faster. The operating question is harder. Which products will be stored there, how will they be replenished, and what happens when one order needs items held in different places?
I would add a warehouse only after a specific customer or capacity problem has been demonstrated and a two-location plan improves the complete outcome. “We are growing” is a reason to investigate capacity. It is not enough to decide that splitting inventory is the right next step.
The alternative may be a better carrier service, a change to the current warehouse, a limited regional trial or placing only selected products in another location. Compare those options before committing the whole catalog.
Name the problem the second site needs to solve
Start with actual orders. Are customers in a particular region consistently receiving a slower service than the website promises? Is the current warehouse unable to process the necessary workload? Is inbound receiving congestion delaying stock availability? Those problems do not necessarily share the same solution.
Separate the time before dispatch from the time after it. If orders wait two days for internal processing, a closer building may not fix the cause. If the carrier journey is the issue, compare routes and services. If the constraint is a product preparation step, assess whether a second site can perform it reliably.
Write the intended improvement in customer and operating terms. For example: “Provide the agreed service to this existing group of regional orders while keeping stock availability and complete cost acceptable.” Avoid making a new location itself the success measure.
Map orders and products together
A destination map shows where customers are. It does not show which products they buy together. Build both views before deciding what inventory to split.
Take a representative period of orders and group them by destination, purchased products and shipping configuration. Include repeat combinations, slower products and relevant seasonal changes. Remove canceled orders or handle them consistently with your purpose, and document the exclusions.
Ask whether a small set of products accounts for the regional demand you want to serve. If so, a limited assortment may be worth testing. If baskets often combine those products with items kept elsewhere, the same plan may generate split shipments.
Do not allocate stock only from total unit demand. Two products with equal sales can have different replenishment times, packaging needs and patterns of being bought together. The allocation should support complete customer orders, not just make each product's distribution look balanced.
Account for the inventory you can no longer share instantly
At one location, available stock can serve any eligible order routed there. At two locations, stock may be in the wrong place for the service you want to offer. A unit shown as available somewhere in the network does not necessarily support a particular local promise.
Decide how the website will determine availability and delivery options. Decide whether an order can be rerouted when its nearest site is short, and what that changes about delivery time or cost. If split shipping is possible, confirm how the customer is informed and charged.
Maersk's inventory planning guidance discusses stock accuracy, demand variation and replenishment timing. My practical recommendation is to apply those questions separately to each proposed location rather than dividing one stock total in half and assuming the original service level survives.
An additional site does not always require duplicating every item. But any plan that holds extra stock needs to explain which uncertainty it covers and how the quantity will be reviewed. “Safety stock at both sites” is not a complete purchasing decision.
Compare the complete scenarios
Use the same order group to compare the current arrangement, an improved single-site arrangement and the proposed two-site arrangement. Price the work each would actually require.
Include inbound transport, receiving, storage, order handling, outbound delivery, transfers between sites, relevant systems and internal coordination. Identify minimum commitments and one-time setup work. Estimate additional working inventory separately from monthly operating cost so the cash requirement remains visible.
Here is a fictional monthly comparison. These invented amounts demonstrate a decision method, not market pricing.
| Category affected by the decision | One site | Two sites |
|---|---|---|
| Outbound transport | $14,000 | $11,800 |
| Storage and receiving | $3,200 | $4,100 |
| Transfers between sites | $0 | $600 |
| Additional systems and coordination | $300 | $650 |
| Total of these categories | $17,500 | $17,150 |
The apparent $2,200 outbound saving becomes $350 after the other listed changes. If setup costs $6,000, dividing that by $350 gives roughly 17 months to recover setup under these assumptions. That simplified calculation excludes financing, taxes, extra inventory cash and any other omitted effects. If the operating saving does not persist, the calculation does not hold.
Now add the customer-service benefit the plan is intended to deliver. A modest cost saving may still be worthwhile if the improvement is important and demonstrated. A cost increase might also be justified, but it should be an explicit choice rather than hidden behind the lower outbound freight line.
Stress the assumptions that could reverse the decision
Test a quieter month, a shifted regional mix and a delay in replenishing one site. Look at what happens when a common basket cannot be filled from a single location.
For each case, calculate the practical response. Does the system send everything from the other site? Split the order? Wait? Offer a different delivery estimate? Moving cost and delay between those options can change whether the second warehouse achieves its purpose.
Ask the provider to explain how promotional capacity is reserved and what information it needs from you. Do not assume unused space means unlimited receiving or packing capacity on the required day.
If a small change in an uncertain assumption reverses the result, use a limited trial to learn more. Do not hide the sensitivity behind a single forecast that looks precise because the spreadsheet has decimals.
Set routing rules your team can explain
Order routing decides which warehouse fulfills an order. A useful rule considers availability and the required service as well as distance. It also needs a clear response when the normal choice fails.
Before launch, test orders with all items in both sites, all items in only one, and items split between sites. Test a stock correction after an order is placed, a cancellation and a failed system update. Confirm that the customer does not receive contradictory dispatch messages.
Support needs to see the decision and the resulting shipments. Finance needs to relate costs to the correct order or transfer. The warehouse teams need a consistent view of what they should pick. A routing feature is ready when those handoffs work, not merely when an algorithm selects the geographically closest site.
The website reliability guide is relevant here: stock and fulfillment connections become part of the customer-facing website, and they need ownership and failure checks after launch.
Pilot the part that has a clear reason to move
Choose a manageable product group and region that represent the intended benefit. Record opening stock and the plan for replenishment. Agree on what evidence would support expansion and what would cause you to pause or revise the trial.
Track the complete customer handoff, stock availability, split orders, transfers and actual cost. Read the exceptions. A promising average can conceal repeated failure for a product combination that matters to the business.
Keep the comparison honest. If the trial runs during a quiet month while the baseline includes a major promotion, explain the difference. If the product range changes, do not attribute every cost movement to the location alone.
Have a practical way to stop. Know how remaining inventory would be reconciled and moved, how outstanding orders would be completed and what commitments remain. A reversible trial is valuable only when the route back has been considered before stock and systems become dependent on the new site.
Expand because the evidence supports it
After the trial, distinguish three outcomes. The location may have worked as intended and be ready for more suitable volume. It may have exposed a fixable routing or stock-allocation problem. Or the expected benefit may not be large enough to support the added complexity.
Each is a useful result. Keeping a smaller assortment at the second site can be a good decision. Returning to one improved operation can also be a good decision. The number of warehouses is not a measure of brand maturity.
I would expand only when the team can explain which orders benefit, how the stock remains available, what the arrangement costs and how exceptions are handled. If those answers remain uncertain, solve them before distributing more inventory.
Questions about multiple warehouses
At what order volume should a brand add a second warehouse?
There is no universal threshold. Destination patterns, product combinations, capacity, inventory availability and service commitments matter alongside volume. Use real orders to compare a specific two-site plan with the current operation and an improved single-site alternative.
Should every product be stocked in both locations?
Not automatically. Consider demand, products bought together, replenishment and handling needs. A limited assortment can reduce commitment, but it must be tested for split shipments and unavailable combinations. Decide from complete order behavior rather than equal stock allocation.
Is a faster carrier journey enough to justify expansion?
It is one possible benefit. Check the complete time from order to usable delivery, along with receiving, stock availability, cost and customer support. A closer site does not help an order that waits for missing stock or is routed incorrectly.
Continue with the DTCUP warehousing guides for related decisions, or use the store launch and migration guide when the change requires a coordinated release across website and fulfillment systems.
Sources
Still have a question?Ask the DTCUP Community ↗


Leave a Reply
You must be logged in to post a comment.