Amazon inventory tracking stops being simple once your operation starts to scale.
In operations with few SKUs and a single logistics flow, inventory usually sits in one location and is controlled by a single system. In that setup, the question “how many units do we have?” usually has a straightforward answer.
Once FBA and FBM start to coexist, inventory is no longer in a single place. Part of it is in Amazon’s fulfillment centers, part is in your own or 3PL warehouse, part is in transit, and part is already committed to orders. Each system records these units in a different way and at different times.
The number that once represented total inventory now becomes a combination of multiple categories, sources, and updates. Without a structured consolidation, there is no longer a single inventory figure. There are several simultaneous positions, depending on which system you are looking at.
When this distinction is not formalized, tracking stops supporting decision‑making. What looks like a simple question, “how many units do we have?”, turns into something more specific: how many units are actually available to sell, on which channel, considering reservations, inbound inventory, and operational constraints.sell.
This is where the Amazon inventory tracking problem begins. It is not about counting; it is about operational definition. And that is what determines whether your inventory is merely monitored or truly controlled when using both FBA and FBM on Amazon.
Quick note:
FBA (Fulfillment by Amazon) means Amazon stores and ships your products from its own fulfillment centers.
FBM (Fulfillment by Merchant) means you store and ship orders yourself or through a 3PL, while Amazon acts only as the sales channel.
The difference between FBA and FBM is not just about logistics. It defines where inventory is controlled, how it is classified, and which system determines what can or cannot be sold.
FBA, Fulfillment by Amazon, means inventory is shipped to Amazon fulfillment centers. Amazon stores, picks, ships, and processes returns. The inventory is physically under Amazon’s control, and visibility depends on the inventory categories and reports available within the platform itself.
FBM, Fulfillment by Merchant, means inventory remains under the brand’s control, whether in its own warehouse or with a 3PL provider. Amazon acts only as a sales channel. Inventory availability depends on the integration between the brand’s internal systems and the marketplace.
The key difference is not just who ships the order. It is which system controls the physical inventory.
With FBA, the main risk lies in misinterpreting Amazon’s internal inventory categories. With FBM, the main risk lies in synchronization between systems.
To understand why this matters, we need to go deeper into what “inventory” actually means inside Amazon.
If the challenge of inventory tracking is an operational definition problem, the first common mistake starts with how FBA itself organizes inventory.
Inside Amazon, inventory is not a single number ready for decision‑making. It is split into different operational statuses, each with distinct implications for sales, replenishment, and risk.
Amazon breaks inventory into distinct categories:
Taken together, these statuses mean that “inventory” inside Amazon is not a single, ready‑to‑use number, but a collection of operational buckets that behave differently. Any serious inventory tracking or replenishment model that involves FBA needs to start by deciding which of these buckets count as truly sellable inventory, which belong to future availability, and which should be treated as operational risk.
Imagine you are selling a product on Amazon. In your Amazon inventory view, you see the following numbers for that SKU:
If someone asks, “How many units do you have?”, you could say 340. But that does not mean you can sell 340 units right now.
You can only sell 100 units now. The 20 units are already promised to customers. The 200 units have not arrived yet. The 15 units cannot be sold because the listing is blocked or inactive. The 5 units are damaged.
That is the core problem: the total number exists, but it does not answer the right question.
The real question is: “How many units can I actually use for this decision right now?”
In other words, the total inventory number is just context. For operational decisions, what really matters is the subset of units that are actually sellable right now for the decision in front of you, whether that is placing a reorder, adjusting prices, launching a campaign, or allocating units across channels.
Now imagine you keep the product in your own warehouse. Your WMS shows 120 units on hand and that number is being sent to Amazon.
Amazon, however, only refreshes the displayed quantity every 15 minutes. During one of those 15‑minute windows, you sell 15 units on your own site and other marketplaces.
Your warehouse system already knows there are only 105 units left, but Amazon is still showing 120 units available to sell.
If more customers place orders on Amazon during this lag, you are effectively selling inventory that no longer exists. That turns into cancellations, late shipments, and avoidable hits to your account health and Buy Box eligibility.
In this case, the problem is not how inventory is categorized. The problem is that inventory updates between systems are delayed, so “available” on Amazon is not the same as “available” in your physical warehouse.
The physical inventory can be correct.
What changes is:
That is why inventory tracking is not just “seeing the number”. It is understanding which number really matters for each decision.
Each category supports different decisions. If inbound inventory is treated as immediately available, your coverage can be overestimated. If reserved inventory is ignored, the risk of overselling goes up. If stranded inventory is not monitored, your apparent availability is artificially inflated.
In FBM, the problem is different but just as relevant. Available inventory depends on synchronization between your WMS, ERP, and Amazon. Even small delays are enough to trigger cancellations or ranking losses, and over time this failure can compromise your entire operation.
In many operations, the problem is not a lack of integration. The data is correct inside each system. The error lies in the implicit definition of inventory position.
Finance might exclude inbound inventory when calculating tied‑up capital. Operations might include inbound when estimating coverage. The marketplace manager might look only at available units. The planner might subtract reserved units. Each number is correct in isolation.
Inventory tracking fails when “inventory position” is not a formal, shared definition. At that point, the number is no longer just a data point; it becomes a decision variable.
This problem gets worse in multichannel environments without alignment, where the organization ends up operating with multiple versions of the truth.
Once you accept that “inventory” is not a single number, the next step is to make a deliberate choice: what exactly counts as available to sell across Amazon and the rest of your channels. A unified available‑to‑sell (ATS) position is simply one clear definition that every system and every team uses as the reference for decisions.
The starting point is simple: how many units physically exist in your network. That includes all locations where you actually store product (FBA, FBM warehouses, 3PLs, retail backrooms, etc.). From there, you decide what to add and what to subtract.
In practical terms, your base should be:
Inbound and transfers are where most Amazon inventory tracking models break. If you treat all inbound as available, you will constantly overestimate coverage.
Define explicit rules, such as:
The next step is to remove everything that is already promised. That includes:
The goal is to make sure that available‑to‑sell inventory reflects what is truly free to be sold again, not what is already spoken for.
In a multichannel operation, you rarely want to expose the entire global ATS number to every channel at once. A unified ATS position is the pool you can work with; then you decide how to allocate it.
Common patterns include:
The important part is that these rules are explicit and encoded in systems, not hidden in spreadsheets or individual habits.
Finally, turn this definition into a formal, shared contract inside the company. That means:
Once this unified available‑to‑sell position exists, Amazon inventory tracking stops being a collection of disconnected reports and becomes a consistent decision engine. You can push harder on Buy Box, plan promotions, and expand channels knowing that every sale is hitting the same underlying reality.
For most brands, the real challenge is not defining this model on a slide, but keeping it live across all channels, systems, and SKUs. That is exactly the kind of problem modern inventory planning platforms like Flieber are built to solve: turning your available‑to‑sell definition into a day‑to‑day operating reality instead of a one‑off spreadsheet exercise.
Up to this point, the problem might sound conceptual. But when this fragmentation is not organized under a single, unified inventory position, it stops being theory and starts showing up in day‑to‑day operations.
When FBA, FBM, and other channels coexist, the failure points multiply:
When these issues are not reconciled into a single, structured view of inventory, each channel and team ends up acting on a different reality. The only way to make consistent decisions across FBA, FBM, and other channels is to define one unified “available to sell” inventory position that every system and stakeholder aligns to.
And if you are at the stage where spreadsheets are no longer enough, it may be time to look at a dedicated inventory planning platform such as Flieber to keep your Amazon and multichannel available‑to‑sell position under control.