Ecommerce Inventory Management: How to Run Weeks of Coverage Across Every Channel

Every brand owner remembers the first time it happened. An email from Seller Central, or a customer asking why they can’t check out. You go and look, and the listing is out of stock — while your ads keep spending, and 4,000 units of that exact product sit in a container, or at a prep center, or in a receiving queue somewhere in Pennsylvania.

You didn’t have an inventory shortage. You had a visibility problem. This guide fixes that. We’ll cover what actually changes when you sell through Amazon FBA, TikTok Shop’s FBT or Walmart WFS, what a 3PL really does for you, and then the weekly twenty-minute routine that ties it all together. No software required — a spreadsheet and the right two numbers will do.

---

In this guide

• Inventory is a coverage problem, not a storage problem
• The five places your inventory hides
• What each platform does to your numbers
• What a 3PL actually is — and why it changes the math
• The routine: count your coverage twice
• How much coverage is enough? The floor and the ceiling
• The number most brands get wrong: not length, but swing
• The bottom line
• Frequently asked questions

Inventory is a coverage problem, not a storage problem

There are only two ways to get inventory wrong. They feel completely different, and they cost completely differently.

Too little: the loud failure

You lose the sale. That part is obvious. What stings is everything that goes with it.

You also lose your search rank, your algorithmic momentum, the buy box, and the review velocity you spent months building. If you were running paid traffic, you kept paying to send people to a page that couldn’t take their money. And when you come back in stock, you don’t return to where you left off — you start climbing again.

That’s why a stockout on your hero product is rarely a one-week problem. It’s a one-quarter problem.

Too much: the quiet failure

This is the one that closes more brands, and nobody posts about it.

Inventory is cash in a physical shape. Every unit on that shelf is money you already spent and can’t spend again — not on ads, not on your next product, not on payroll. Ask anyone who has watched a payroll run approach while their working capital sat in a warehouse in boxes. It is one of the quieter operational bottlenecks that stalls a growing brand.

Holding it isn’t free either. Planners usually budget 20–30% of inventory value per year once you add up storage, tied-up capital, damage and obsolescence. Stock that sits for a year has quietly eaten a quarter of its own value.

So what should you actually track?

Here is the problem with the number on your dashboard. Units on hand tells you almost nothing. Is 4,000 units good? It depends entirely on how fast you sell. For one SKU that’s a comfortable quarter. For another it’s eighteen months of dead stock.

The number only becomes useful when you turn it into time:

Why time rather than units? Because time is the only unit you can compare against a lead time.

Hold “4,000 units” up against “eight weeks at the factory” and you learn nothing; they aren’t the same kind of thing. Hold “seven weeks of coverage” up against “a ten-week reorder cycle” and the answer is instant: you are going to run out. Every decision in this guide comes out of that single comparison.

The five places your inventory hides

Before you can count coverage, you need to know what you’re counting — and this is where most brands lose the thread.

If you sell on more than one channel, the same SKU exists in five states at the same time. Only one of them can take an order today.

1. Sellable at a channel. Checked in and live at FBA, FBT, WFS or your 3PL. This is the only stock that can convert a visitor right now.
2. Received but not sellable. Physically in the building, but stuck — in a receiving queue, in reserved or transfer status, waiting on prep or labeling.
3. In transit. On a truck, in a container, in a parcel headed for a fulfillment center.
4. In production. The purchase order sitting with your manufacturer.
5. Gone. Damaged, returned and not restockable, quarantined, or held pending a quality call.

Now here’s the uncomfortable part — look at where each of those numbers actually lives.

State 1 is on every dashboard you own. Part of state 2 shows up if you know which report to open. State 3 lives with your freight forwarder. And state 4 — often the biggest pile of all — lives in an email thread with your manufacturer. That is exactly why it never makes it into anyone’s inventory view, and exactly why brands get blindsided.

Amazon already counts your pipeline. You should too.

If you need a reason to take this seriously, take it from Amazon.

Your FBA capacity limit isn’t based on what’s in the warehouse. Amazon calculates it from “inventory on-hand in Amazon’s fulfillment centers and shipments sellers have created that have not yet arrived.” Units you haven’t even handed to a carrier are already eating your allowance. (Amazon’s own announcement spells this out.)

The platform is counting in two places. Most sellers are counting in one.

One last thing while we’re here: every SKU you add multiplies this bookkeeping by the number of channels you sell on. Five SKUs across four channels is twenty coverage numbers to keep straight. A tight product range isn’t only a brand decision — it’s an operational one, which is why disciplined brands run fewer SKUs than they could.

What each platform does to your numbers

Every fulfillment channel does the same basic job: hold your stock, ship your orders. But each one bends your inventory math in its own way, and the differences are the sort of thing you usually learn the expensive way.

A note on what follows: this is about time and rules, not fees. Rate cards change every few months and are the wrong thing to build a supply chain around. What matters for planning is how long each channel takes to make your stock sellable, and who controls that.

Amazon FBA (Fulfillment by Amazon)

Four things about FBA will show up in your planning whether you account for them or not.

• Delivered is not sellable. There’s a receiving and check-in step between your truck arriving and your units going live. It stretches when the network is busy — and it is busiest exactly when you need it most, in the fourth quarter.
• Your capacity is capped monthly, and measured in cubic feet. Amazon sets one limit per account, announced in the third full week of each month in the Capacity Monitor. It measures usage “in cubic feet (vs. number of units), which better represents the capacity” — so a bulky SKU eats your allowance far faster than a dense one, whatever the unit count says.
• The allowance depends on how well you manage stock. Amazon says limits “are influenced by sellers’ IPI scores, as well as other factors such as sales forecasts for their ASINs, shipment lead time, and fulfillment center capacity.” Think of the Inventory Performance Index as a credit score for your inventory; Amazon points to 400 as the level that keeps you healthy. Notice shipment lead time in that list — slow, unpredictable resupply costs you allowance, not just sales.
• One purchase order can go live on several different dates. Amazon usually wants your inbound shipment spread across several fulfillment centers. You can consolidate into fewer, but that carries a per-unit placement charge, so brands trade one against the other. Either way, one PO can become four arrivals checking in on four different days. Never treat an FBA quantity as a single arrival date.

TikTok Shop and FBT (Fulfilled by TikTok)

Stock gets in one of two ways: pallet freight against a booked delivery appointment, or small parcels logged with a tracking ID. You can ship to an East or West hub warehouse that then spreads your units across fulfillment centers, or send straight to a named center.

Two published terms matter for planning. TikTok gives 60 days of free storage on each inbound shipment — which quietly rewards smaller, more frequent shipments over one big drop. And TikTok says 98% of FBT orders ship within one calendar day. The outbound leg is fast, so your risk sits on the way in, not the way out.

But the mechanics aren’t the real story with TikTok. The volatility is.

Today TikTok lists three routes — FBT, TikTok Shipping (their labels, your warehouse) and Seller Shipping (your carriers, your warehouse) — available “depending on your business eligibility”. But in late January 2026, US sellers were told they’d have to move off independent shipping onto FBT or TikTok-integrated logistics, starting in late February and finishing by the end of March. Then on February 17, after loud seller pushback, TikTok emailed everyone that “Seller Shipping remains unchanged, and previously shared deadlines are not going into effect.” (Modern Retail covered the reversal.)

Read that timeline again as an operator. A channel told brands to rebuild their fulfillment in five weeks, then reversed it in three. The mandate is paused, not canceled.

Walmart WFS (Walmart Fulfillment Services)

WFS is the most honest of the three about the step that matters most to planners. Walmart states it “normally takes up to 2 business days to receive inventory from the time it was delivered.” But items needing unplanned prep work, or arriving during holidays or peak season — which Walmart defines as October 1 to December 31 — can take up to 10 business days. You also have to create the inbound order in Seller Center before you ship, and ship to every fulfillment center in the plan.

Sit with that range for a second, because it’s the most instructive number in this article. Two days to ten days isn’t a long wait. It’s a wildly inconsistent one — the same step can take five times longer depending on the week and how clean your paperwork was.

Hold onto that. Inconsistency, not length, is what forces you to carry extra stock, and we’ll put a number on it in the last section. Walmart even builds the uncertainty into its own process: you have to wait at least ten days from delivery before you can open a receiving dispute.

The one-page comparison

If you remember nothing else from this section, remember this table:

Channel	Fills orders for	Time from delivered to sellable	Who controls it
Amazon FBA	Amazon only	Receiving + check-in; longer in Q4	Amazon
TikTok FBT	TikTok Shop only	Hub or direct inbound; no published SLA	TikTok
Walmart WFS	Walmart only	Up to 2 business days; up to 10 in peak	Walmart
Your 3PL	Anything you point it at	Hours to a few days, by provider	You

What a 3PL actually is — and why it changes the math

That last row in the table deserves its own section, because a 3PL is the piece most growing brands understand least and benefit from most.

A typical 3PL takes in your inbound freight, stores it, picks and packs orders from your store, ships them, and processes returns. Most also do value-added work: kitting, bundling, inserts, relabeling. And many will prep and forward stock into FBA or WFS for you.

That last one is the whole ballgame. Here’s the distinction worth tattooing somewhere:

• A platform warehouse is an endpoint. Stock in FBA fills Amazon orders. Stock in FBT fills TikTok orders. Stock in WFS fills Walmart orders. Each pool is walled off from the others.
• A 3PL is a hub. The same pool can fill your own store, wholesale purchase orders, marketplace orders you ship yourself — and top up FBA, FBT or WFS when any of them runs low.

Which is why most brands, past a certain size, end up in the same shape: a 3PL in the middle, feeding the platform warehouses as they draw down.

It costs a bit more per unit than shipping factory-direct into Amazon. What you buy for that premium is two things worth having — the ability to move stock to whichever channel is outrunning the others, and a fulfillment path that survives a platform changing its rules on you.

How to choose one (the questions nobody asks)

When brands shop for a 3PL they compare pick fees. Reasonable, but it isn’t what will hurt you.

Ask instead for average receiving time and how much it varies — in writing. Big automated operations check stock in fast and predictably. Smaller ones are often cheaper and more flexible, but slower and much less consistent. You are buying predictability, and you should make them quote it.

One more trap. If you split inventory across multiple 3PL locations to shorten delivery times, you have just multiplied your planning work. Every location needs its own coverage number, because a healthy national total tells you nothing about the East Coast running dry on Thursday.

The routine: count your coverage twice

Everything so far has been groundwork. This is the part you actually do.

It replaces most inventory software. Once a week, same day, per SKU, per channel. Twenty minutes in a spreadsheet once the inputs are wired up. Here are the six steps.

Step 1 — Get your run rate

Average units sold per week, per channel, over the last four weeks.

One important exception. If the SKU is growing, don’t use the trailing average — use the trend. You’re buying stock that lands weeks from now, so what matters is the demand you’ll have when it arrives, not what you had last month. Planning a fast-growing SKU off a backward-looking average is the single most common way brands under-order their own winner.

Step 2 — Coverage at each channel, one at a time

Sellable units at that channel ÷ that channel’s run rate. This is the number you already track. It tells you how long that channel keeps selling if nothing else ever arrives.

Do it per channel, and never average across them. Section 4 explained why: these pools are walled off. Units in FBA cannot fill a TikTok order. Add all your sellable stock together, divide by total demand, and you get a comfortable-looking blended number that can completely hide a channel about to go dark. The number that matters is your thinnest channel, not your average one.

Step 3 — Coverage in progress

Now the number almost nobody tracks: everything on order but not yet sellable. In production, in transit, in receiving.

Units already earmarked for a channel count toward that channel. A production batch you haven’t split yet is a shared pool — measure it against your total weekly demand and decide the split when you allocate it.

Getting this number is less about math than about admin: ask your manufacturer for confirmed quantities and dates, then write them somewhere your other numbers live. If it only exists in your inbox, it doesn’t exist.

Step 4 — Compare the total against your reorder cycle

Add sellable and in-progress together. Then compare that total against how long a full replenishment genuinely takes, end to end:

If your total coverage is less than your replenishment cycle, you are going to stock out. The order you place today cannot arrive before the shelf empties. That isn’t a forecast — it’s arithmetic.

Step 5 — A worked example

Let’s put real numbers on it. A serum selling 5,000 units a week across three channels — 3,000 on Amazon, 1,000 on TikTok Shop, 1,000 through the brand’s own store. The replenishment cycle is ten weeks: eight weeks production, one week transit, one week receiving. The purchase order currently in production was placed three weeks ago, so it has five weeks left on the line.

First, sellable stock — one channel at a time:

Channel	Sellable units	Sells per week	Weeks of coverage
Amazon FBA	24,000	3,000	8.0
TikTok Shop (FBT)	6,000	1,000	6.0
Own store (via the 3PL)	10,000	1,000	10.0

Then the pipeline. Nothing in transit, and 25,000 units in production — measured against the brand’s total 5,000 a week, because that batch hasn’t been split between channels yet, so it’s 5.0 weeks. Sellable plus pipeline is 65,000 units, or 13.0 weeks of total coverage against a ten-week cycle. Three weeks of slack. Looks fine.

It isn’t fine. Look at the thinnest channel.

TikTok has 6.0 weeks. The batch in production becomes sellable in seven weeks — five to finish, one in transit, one in receiving. Six is less than seven. Even in a calm week, before anything has gone wrong, TikTok is already on track to run dry about a week before resupply lands. A blended figure would have reported 8.0 weeks of sellable coverage and shown you nothing at all.

Now let one thing change — the thing that actually happens. A creator video lands and TikTok’s run rate jumps from 1,000 to 2,500 a week. Not a single unit has moved. No purchase order has changed. But:

• Amazon is untouched at 8.0 weeks. Its demand didn’t change, so its coverage didn’t either. This is exactly why you don’t average.
• TikTok collapses to 2.4 weeks against seven weeks until resupply — short by roughly four and a half weeks.
• Total coverage falls to exactly 10.0 weeks, which is the replenishment cycle itself. The brand is sitting precisely on the floor, with nothing left over for anything else going wrong.

Nothing at the factory can fix a four-and-a-half-week hole — that batch is still five weeks from finished. The only lever faster than the problem is the 3PL, which can push units into FBT in days rather than weeks. It holds 10,000, which covers most of the gap but not all of it, and only by borrowing from the brand’s own store.

And that’s the real lesson. Moving stock buys you weeks. It doesn’t replace a purchase order. The signal that mattered was sitting there in the calm week, in two numbers: TikTok’s six weeks against a seven-week resupply, and only three weeks of slack over the cycle. Both were readable before the video ever landed.

Step 6 — Set your triggers, then keep running it

You need two triggers, not one:

1. Place the production order when total coverage drops toward your replenishment cycle plus your buffer. Not when a platform looks thin — by then the decision has been made for you.
2. Move stock between channels when your thinnest channel’s coverage falls below the time it takes to resupply that channel. This is the shorter clock, and it usually fires first.

Then re-run the whole thing weekly. Not out of diligence — because the denominator moves. Coverage is a ratio, and a SKU can go from comfortable to critical without anyone touching a single unit, purely because it started selling faster. That’s the good problem. It’s also the one that catches brands out.

The weekly checklist

Copy this into your spreadsheet and work down it once a week:

• Run rate per channel — last four weeks, or the trend if the SKU is growing
• Sellable units per channel → coverage per channel
• Which channel is thinnest? Is it below its own resupply time?
• Units in production and in transit → pipeline coverage
• Total coverage vs. replenishment cycle — how much slack?
• Anything to reorder? Anything to move between channels?

How much coverage is enough? The floor and the ceiling

“How much stock should I hold?” is the question every brand owner asks, and the honest answer is that there’s no universal number — but there is a range, and it has two edges.

The floor: your reorder cycle plus a buffer

Coverage above your replenishment cycle isn’t a target. It’s the bare minimum required not to run out, and it gives you zero room for anything going wrong. On top of that you need a buffer.

Two things size that buffer, and neither is a rule of thumb:

• How much your sales swing week to week. A steady repeat-purchase SKU needs far less cushion than one that lives and dies on creator posts.
• How much your supply swings. This is the one brands consistently underestimate, and it gets the next section to itself.

Not every SKU deserves the same protection

Here’s a classic technique that costs nothing to adopt and saves a lot of cash.

Your two or three heroes carry the rank, the reviews and the repeat rate — protect them aggressively. The long tail does not deserve the same treatment, and funding it as though it does is precisely how brands end up cash-poor and stocked-out at the same time.

Related idea worth naming: your service level is simply how often you’re willing to run out. Aiming to fill 99% of demand from stock costs dramatically more buffer than 95%. That’s a business decision, and it should be a different one for an A SKU than for a C.

The ceiling: order what you can sell through

Now the other edge — and this is the one we’d put in front of any brand placing its first few production orders.

Order what you’re confident you can sell through in three to twelve months.

Notice what that isn’t. It isn’t “buy as little as possible.” That advice has its own failure mode, and it’s a painful one: stocking out of the product that was finally working. Starting small is often a consequence of this rule for an unproven SKU, not the rule itself.

The discipline is sell-through. Stock that will take two years to move has frozen cash you need now, is accruing storage fees, and in beauty and personal care is aging against a shelf-life and period-after-opening clock. If you’re sizing a first production run, this is the same logic we walk through in the 2026 launch guide.

When your MOQ breaks the ceiling

This is where the rule collides with manufacturing reality, and where a lot of brands quietly accept a bad deal.

If the smallest batch your manufacturer will run is more than twelve months of demand at your current velocity, you aren’t choosing to hold stale inventory. You’re being pushed into it by a minimum order quantity.

It is worth knowing what is actually achievable before you accept a number. Minimums vary enormously by route: a private label program built on existing formulas can start in the low hundreds of units, while a bespoke formulation carries a higher minimum because someone has to develop it first. If your MOQ is forcing two years of stock onto your balance sheet, the route may be the problem rather than the quantity.

Treat that as a signal, not a fact of life. You have four real options: negotiate the MOQ, split the run across scheduled releases, hold the launch until your velocity justifies the batch, or decide the SKU isn’t worth carrying.

What you shouldn’t do is absorb it silently and call it the cost of doing business. An MOQ is a commercial term. Commercial terms are properties of the supplier you picked — not laws of physics.

The number most brands get wrong: not length, but swing

We’ve saved the most valuable idea in this guide for last, and it’s one that almost never comes up when brands evaluate a manufacturer.

Ask a brand its lead time and you get one number. “Eight weeks.” That single number is hiding the variable that actually decides how much inventory you have to own.

Two factories, same average, very different cost

Picture two manufacturers who both average eight weeks.

The first delivers between seven and nine weeks, every single time. The second ranges from five to fourteen. On paper they’re identical. In practice the second one is dramatically more expensive to work with — and the cost never shows up on an invoice.

Why? Because your buffer has to cover the bad case, not the average one. If a delivery might take fourteen weeks, you have to carry enough stock for fourteen weeks, no matter what the average says. Reliability isn’t a nicety. It’s a line item you pay in frozen cash.

What the swing actually costs

The size of the effect surprises people, so let’s put a number on it.

Take a SKU selling around twenty units a day, with a forty-five-day average lead time, and a target of filling 98% of demand from stock. With a swing of roughly ±12 days, the buffer you need works out around 510 units.

Now halve the swing to ±6 days. Nothing else changes — not the average lead time, not the price, not the demand. The same 98% service level now needs about 280 units.

That is a 45% cut in buffer stock from consistency alone. Multiply 230 units by your unit cost, then by every SKU in your catalog, and you’re looking at the real price of an unpredictable supply chain: working capital parked in a warehouse purely to absorb someone else’s inconsistency.

It compounds in a place you might not expect, too. Remember that Amazon lists shipment lead time among the inputs to your FBA capacity limit. Unreliable resupply doesn’t only force you to hold more stock — it can shrink the allowance you’re given to hold it in.

How to actually get this number

Nobody is going to hand it to you, so measure it yourself. It takes about five minutes per order.

• For every purchase order, write down the promised delivery date and the actual one.
• After five or six orders you’ll have a real average — and, far more useful, a real spread.
• Do the same for each inbound leg: delivered-to-sellable at FBA, at FBT, at WFS, at your 3PL.

Then treat that spread as something to fix, not something to endure. Sometimes it’s a conversation about scheduling and component procurement with your current manufacturer. Sometimes it’s a different manufacturer, or a different route entirely — manufacturing from an existing private label formula carries a materially shorter lead time than developing one from scratch, because the formulation and testing are already done. Either way, the quoted lead time was never the number that mattered — and if you want the view from the other side of that relationship, that’s the side we work on.

The bottom line

Inventory management for a growing brand isn’t really about warehouses, and it definitely isn’t about software. It’s one weekly habit. Turn every pile of stock into weeks of coverage. Count it in both places — sellable at each channel, and in progress with your manufacturer. Compare the total against how long a full reorder genuinely takes. Do that consistently and both failures we opened with stop being surprises: you’ll see the stockout coming while there’s still time to place the order, and you’ll stop funding stock you won’t sell for two years. Start this week, with your top three SKUs and a spreadsheet. The routine matters far more than the tooling.

Frequently asked questions

What is weeks of coverage in inventory management?

Weeks of coverage is how long your current stock lasts at your current sales rate: units divided by average units sold per week. It’s also called weeks of supply, or days of cover for a finer grain. Its value is that it’s expressed in the same unit as your lead time, so you can compare what you hold against how long resupply takes.

What is the difference between FBA and FBT?

They’re the same idea on two different platforms. Fulfillment by Amazon (FBA) stores and ships inventory for Amazon orders; Fulfilled by TikTok (FBT) does the same for TikTok Shop orders. Neither pool can fill the other platform’s orders, so stock committed to one isn’t available to the other. That’s why brands selling on both usually keep a separate buffer that can feed either.

Do I still need a 3PL if I use FBA?

Often yes, once you sell on more than one channel. FBA only fills Amazon orders, so it can’t serve your own store, your wholesale orders, or another marketplace. A 3PL acts as a central hub that can fill those directly and top up FBA or WFS as they draw down — and it gives you a fulfillment path that survives a platform changing its rules.

How many weeks of inventory should an ecommerce brand hold?

There’s no universal number, and any article that gives you one is guessing. Your floor is your own replenishment cycle — production plus transit plus inbound receiving — and you need coverage above that, plus a buffer sized to how much your sales and your supply actually swing. Your ceiling is sell-through: don’t order more than you’re confident you can sell in three to twelve months.

How long does private label manufacturing take?

It depends on the route, and the gap is wide enough to change your whole inventory plan. Manufacturing from an existing private label formula is the fast path, because the formulation and stability testing are already finished — FormuNova’s typical lead time is 4 to 6 weeks. A custom formulation runs 6 to 12 weeks for repeat production once the development project is complete, and development itself is a separate one-time phase before that. Remember that whichever number applies to you is only the first leg of your replenishment cycle — add transit and inbound receiving to get the clock you are actually racing. Our published private label lead times and minimums spell both routes out.

Does a shorter lead time really reduce how much stock I need?

Yes, and consistency matters even more than speed. A shorter cycle lowers your floor directly, because the floor is your replenishment cycle. But a predictable cycle lowers your buffer on top of that, because the buffer has to cover the worst case rather than the average. In the example in Section 7, halving the swing in delivery dates — with no change to the average lead time at all — cut the buffer needed by about 45%.

What is the fastest way to fix a stockout on one channel?

Move stock from a hub you control, if you have one. A 3PL can usually push units into FBA, FBT or WFS in days, where a new production run takes weeks or months. That’s the main practical reason to keep a central buffer rather than sending every batch factory-direct to a single platform. It buys time — it does not replace the purchase order.

---

FormuNova — product development and contract manufacturing for beauty, skincare, hair care and personal care brands. Stone Mountain, Georgia. formunova.com