Multichannel Inventory Management: One Number for Every Channel

Two orders can claim the last unit in the same minute, one on a marketplace and one on your own site. If both listings still show the count from this morning, you have sold stock you no longer have. Multichannel inventory management is how you keep one available quantity and send that quantity to every place you sell.

The painful part is not the definition. It is the canceled order, the stranded cart, and the cash sitting in a bin nobody can see.

What is multichannel inventory management

Multichannel inventory management keeps one available quantity for every place you sell, then updates that quantity when a sale, a receipt, a transfer, or a return changes it. The channels share the number. They do not each keep a private copy that you reconcile after the warehouse closes.

A channel is anywhere a customer can buy: your website, Amazon, eBay, Walmart, Shopify, a wholesale portal, a store counter. The inventory record is not those storefronts. It is the stock you can actually ship, by location, after you subtract what is already promised.

That split is the whole job. On-hand is what the bin holds. Available to sell is what a new order is allowed to take. Publish the first number and you will oversell the second.

Why a second sales channel breaks one spreadsheet

One store and one stockroom can live in a sheet. You see the orders. You can fix a cell before lunch.

A second channel changes the clock. Orders arrive while you are in a pick, on a call, or asleep. The sheet still says 14 because nobody typed the marketplace sale yet. The website still says 14 because it copied the sheet this morning and nobody has touched it since.

I would not throw out a spreadsheet on day one of a second channel. I would throw it out the first week two channels sell the same SKU faster than a person can type. Past that, the sheet is a diary of what you used to have.

What overselling looks like when two orders land together

Overselling is not a reporting problem you notice at month end. It is two confirmations for one unit.

Here is an example, not a count from your warehouse. Both the marketplace listing and your website still show one unit.

The marketplace sells it. A minute later your own site sells it too. You can ship only one of those orders, and the other becomes a cancel plus a refund you did not plan to write.

Marketplaces notice cancels. Your own site notices them too, as a support ticket and a customer who does not come back. The quantity was wrong before either email went out.

The fix is reservation, not a faster evening edit. When the first order imports, available quantity drops before the second channel can sell it. Amazon's listing guidance makes a related point for seller-fulfilled stock: the last quantity you submitted can still sit in the listing attributes after the item sells, while the live purchasable quantity is the one customers can buy. A feed that "went through" is not proof the channel is showing what your bin holds.

What a stockout does to the order you could have shipped

A stockout is the opposite mistake. You had the unit in another building, or in a transfer, or on a store shelf, and the listing said zero.

You lose the sale. You also train the channel to expect that SKU to go dark. If you restock next week and forget to push the receipt, the listing stays at zero while the pallet sits on the dock.

Location is the detail people skip. Shopify's multi-location inventory guide says each location keeps its own count, those counts are not pooled, and a product is sellable only where it is active. Twenty units in a Toronto store do not fill a Dallas web order unless your routing says they do.

Why extra units in the warehouse are a cash problem

Overselling gets the attention. The quiet problem is stock you bought because every channel looked low at the same time.

Three listings each show "reorder soon" against the same 40 units. Purchasing orders another 40. You now own 80, and the original 40 were never in danger. Carrying that extra stock is cash you cannot spend on the SKU that is actually running out.

Overstock is a visibility problem before it is a buying problem. If you can see on-hand, reserved, and incoming by location, the reorder looks different. If each channel only shows its own fear of a stockout, buying follows the fear.

Why a return does not always become available stock

A refund is not a unit back on the shelf. The customer still has it, or the carrier does, or it is in a cage waiting on inspection.

If your available quantity jumps the moment the refund posts, you can sell a unit that is still in a truck. If it never jumps after the unit is inspected and restocked, you hide good stock.

Put the restock on the warehouse event, not on the refund button. Received and sellable are the moment the channel should hear about it. Damaged is a different bucket, and it should not publish.

How a sale changes the quantity every channel can sell

1. Record on-hand by location before you publish a quantity to any channel.
2. Subtract open orders and reservations so available to sell is what a new buyer can take.
3. Hold safety stock out of the published number when you want a buffer for a late receipt.
4. Push that available quantity after a sale, a receipt, a transfer, or a restocked return.
5. Read the live purchasable quantity on the channel, not only the last number your update submitted.
6. Route the order to a location that has the unit, then pick and pack from that location.

A failed push is part of the sequence, not a footnote. If the warehouse moved and the listing did not, you are back to two numbers. Someone has to see the failed update and send it again. Hoping the next order will fix the gap is how the gap lasts all week.

That loop is multichannel inventory management in practice. A nightly file is a later copy of a number that already changed.

What available-to-sell inventory means

On-hand minus promises, then minus any buffer you refuse to sell. That is available to sell. It is the only quantity a new listing should receive.

Here is an example, not your stock. One bin starts the morning holding 20 units.

An Amazon order then reserves 2 of them, the website sells 3, and a wholesale order reserves 4. Available to sell is 11. Hold 2 back as safety stock and the quantity you publish drops to 9.

Publish 20 and every channel can still sell units those three orders already claimed. Publish 11 and you protect the orders but not the late truck. Publish 9 and you keep a small buffer. The right choice is a rule you write down, not a guess at checkout.

Safety stock is not a percent you copy from a blog. It is units you choose because a receipt is often late, or because one channel spikes on weekends. Label it as your rule. Do not pretend a formula invented the number.

How to add a channel without a second stock count

The benefit people actually want is simple. You open Walmart, or a second Shopify store, or a wholesale price list, and you do not hire someone to retype quantities at night.

A new listing should read the same available quantity as the ones you already run. A sale there drops the shared number. A sale on the old channel drops it too. You added a storefront, not a second warehouse in a spreadsheet tab.

This breaks if the new channel cannot accept a quantity update, or if you map the wrong SKU. A pretty connection that syncs the title and ignores stock is not inventory control. Check the quantity field before you celebrate the product feed.

How one warehouse view stops a transfer from going blind

Two buildings feel like two companies when each channel is tied to one of them by habit. Building A looks empty on Amazon because that listing points at A, while B is full and the website does not know.

A transfer is the operational fix. The inventory fix is seeing both on-hand counts before you promise a ship-from location.

Amazon's multi-location inventory program is aimed at Fulfilled by Merchant sellers in the US who share location-level quantities so delivery dates can reflect where the unit actually sits. Amazon says the program is not available for FBA inventory, and that shipping templates are currently limited to 10 locations. Shipping Settings Automation has to be on.

FBA is a different pile. Those units sit in Amazon's buildings. Adding them to your own bin count, without a feed that keeps the two piles separate, makes both numbers wrong. That is a judgment about the math, and it matches Amazon's line that multi-location inventory does not cover FBA.

When a reorder rule is worth setting up

Forecasting earns a place after the available quantity is trustworthy. A model on top of a stale listing just buys the wrong goods faster.

Start with a reorder point and a safety buffer you can explain. Historical sales, including the seasonal weeks you already lived through, are enough to ask "what did we ship last October?" Incoming purchase orders belong in that picture so you do not reorder a truck that is already on the highway.

An alert for low stock and an alert for too much stock answer different questions. Low stock protects the next order. Too much stock protects cash. If your screen only shouts when you are about to stock out, purchasing will keep overcorrecting.

How inventory software differs from a warehouse system

Selling channels need a quantity. The floor needs a place to pick. Those are related, and they are not the same screen.

Question	Shared inventory record	Warehouse system
What can a new customer buy?	Available quantity by channel, after reservations	Not its main question
Where is the unit?	Location on-hand is the input	Bin, pick path, pack, ship
What happens after a return is inspected?	Sellable quantity can publish again	Restock is a floor task
What happens to the order?	The quantity drops when the order is real	Pick and pack complete the promise

You can outgrow a store admin and still not need a full warehouse build. You can also have a beautiful available quantity and a floor that cannot find the bin. If both problems are yours, connect them. Do not buy a pick-path tool and hope it will stop Amazon from showing yesterday's count.

What to check before you retire the spreadsheet

A sheet is still honest when one person can see every order before another customer checks out. Write down the moment that stops being true.

What you run	Stay on the sheet a while longer	Move the quantity into one record
One shop, one room, few orders a day	You can edit before the next sale
Same SKU on two channels		The second sale cannot wait for a nightly paste
More than one ship-from site		On-hand in one building is not a promise for the other
FBA plus your own bins		Keep the piles separate, or both counts lie
Wholesale orders that reserve a pallet		A reserved pallet is not available to the website

Objections come up here, and they are fair. You can keep the channels you already sell on. The listings stay on those marketplaces. What moves is the quantity behind them.

You can keep the warehouse you have. Routing still has to name a building that holds the unit. Migration is the cutover of that quantity, not a new brand of shelf.

B2B and the website can share stock if reserved wholesale units leave the available number. If they do not, the website will sell the pallet.

Order volume matters only in the sense that a manual edit does not survive a busy hour. There is no magic order count where software becomes mandatory. There is a morning when two confirmations exist for one barcode.

Questions operators ask next

Does each channel need its own inventory count?

No. Each channel needs the available quantity it is allowed to sell. A private count per listing is how two channels sell the same unit.

Is on-hand the same number you should publish?

On-hand is what the location holds. Available to sell subtracts orders you already promised, and any safety stock you refuse to sell.

Can FBA units and warehouse units be one total?

Do not add them into one sellable pile. FBA sits in Amazon's network, and Amazon's multi-location inventory program does not cover FBA. Keep that quantity identifiable on its own.

How is multichannel different from omnichannel?

Multichannel means you sell in more than one place. Omnichannel means a shopper can start in one place and finish in another, so the same available quantity has to be true for the store and the site. The inventory math is shared either way.

When is a spreadsheet still enough?

While one person can update the quantity before the next order. The sheet stops being enough when two channels can sell the same SKU before anyone edits the cell.

Where this sits in the rest of the operation

The quantity is the start, not the whole operation. An order has to land somewhere, a building has to pick it, and a receipt or a restock has to raise the number again. If those steps live in separate tools that do not share the available quantity, you rebuilt the spreadsheet with extra logins.

Inventory management in OneChannelAdmin syncs stock for Amazon, eBay, Walmart, and Shopify, drops available stock when orders import, and raises it when a purchase order or a transfer completes. It keeps on-hand and reserved quantities by warehouse location, and it flags a failed channel update so the update can be retried. The retry matters, because a push does fail.

The Amazon connection describes the same handoff for that channel: a sale reduces available inventory so another channel cannot take the unit, then the order can route to a warehouse, a store, or a 3PL. Order and fulfillment management is where that routing lives, including marketplace and wholesale orders in one queue. Warehouse management is the floor side, from the pick through a restock after a return.

See how inventory stays in one place on the inventory management page. That is the next look if the quantity problem in this article is the one you are actually living with. OneChannelAdmin will not invent a savings percent for you, and a sync log is still something a person should watch.

Sources

Shopify's help page on multiple locations says each location has its own count, those counts are not pooled, and a product sells only where it is active.

Amazon's multi-location inventory page covers Fulfilled by Merchant sellers in the US, excludes FBA, requires Shipping Settings Automation, and currently limits shipping templates to 10 locations.

The Selling Partner API note on seller-fulfilled listings shows that the last quantity submitted and the live purchasable quantity are not always the same number.

Product behavior for inventory, the Amazon connection, orders, and the warehouse comes from the live pages linked in the previous section.

OneChannelAdmin Team writes about commerce operations.