Inventory Planning

When should I reorder this ASIN?

Forecasting FBA reorder dates from real sales velocity — and why the honest answer is a range with a deadline, not a date.

AmazeBase 12 min read Inventory & Demand

Stock running down at sales velocity into a reorder range with a deadline

Knowing how many units you have in FBA is easy. Knowing when you'll need to reorder is the hard part — and the gap between the two is sales velocity.

Podcast episode 21 minutes 37 seconds

This audio is AI-generated with NotebookLM from this article.

A thousand units means nothing on its own. At 10 a day it's a hundred days of cover. At 50 a day it's twenty. And if velocity moves from 10 to 50, the plan you made last month isn't slightly wrong — it's about a different business.

So a reorder forecast isn't a number you calculate once. It's a standing estimate that should move every time the evidence moves.

One caveat, stated once and not repeated: a forecasted reorder date is not a prediction. It's the output of a calculation over assumptions you can name. When an assumption changes, the date changes. That's the system working, not failing.

Here's how to build the estimate, and — more usefully — how to read it.

The worked example — one SKU, carried through the whole article

SKU
AB-2210
Sellable units in FBA
1,240
Inbound units
800
Inbound expected
2 Oct
Last day the sales data covers
5 Sep
Supplier lead times, last 5 orders
48 · 53 · 59 · 67 · 72 d

01Two dates, not one

Sellers ask "when do I run out?" The system has to answer a harder question first.

  • The stockout date is when usable inventory reaches zero under a stated demand assumption.
  • The order-by date is the last day you can place the order and still have units land before that happens.

They are separated by your entire replenishment lead time plus whatever buffer you've decided to carry. With a 60-day supplier, an ASIN with 45 days of cover isn't comfortable — it's already fifteen days late.

Waiting for inventory to look low is a strategy that only works if your supplier is faster than your customers.

Every section below exists to make one of those two dates more honest.

02The window problem, and how to actually resolve it

Sales velocity is units sold ÷ days. The arithmetic is trivial. The choice of days is the whole decision.

AB-2210, trailing windows ending 5 Sep. Every window is a defensible number; they disagree by 57%.
WindowUnitsPer day 
90 days2,70030.0
60 days2,04034.0
30 days1,20040.0
7 days33047.1
Weighted—44.3

The ladder is monotonic: every shorter window is faster than the one before it. That shape is the signal — not any single row. It says demand is accelerating, and it says so more convincingly than a 7-day figure ever could on its own, because a spike shows up in one window while a trend shows up in all of them.

A default you can defend

Blend the responsive window with the stable one and lean recent:

weighted velocity = 0.6 × (last 7 days)
                  + 0.4 × (last 30 days)

                  = 0.6 × 47.1 + 0.4 × 40.0
                  = 44.3 units/day

This reacts to a real change within about a week, and it can't be dragged around by a single unusual day the way a raw 7-day figure can. Use it as the default and override it deliberately.

When to override it

Discount the short window when you can name the cause and the cause has an end date. A coupon that runs until Friday, a competitor who'll be back in stock in two weeks, a Prime Day tail — these are events, not demand. Reprice the forecast to the pre-event baseline and note when to re-check.

Trust the short window when the ladder is monotonic and the gap has persisted for two consecutive weeks. A step change with no nameable cause is usually a real change in rank or competitive position.

The rule that matters most

Use the fast velocity for the date. Use the slow velocity for the quantity.

The two errors are not symmetrical. Acting early costs you a few days of carrying cost and some optionality. Ordering too much costs you cash you can't get back for months. So let the optimistic number decide when you look, and the conservative number decide how much you commit.

03Anchor to the data, not to today

Here is the failure that quietly breaks more reorder forecasts than any modelling choice.

Your sales data ends on some date. It's rarely today — reports settle late, imports run on a schedule, someone forgot. If your last transaction data covers through 5 September and your system computes "last 30 days" against today's date, it averages real sales with days that have no data in them yet.

Twelve days of lag against a 30-day window understates velocity by roughly 40%. Divide your stock by that and the system reports comfortable cover on a SKU that's about to go dark — in green, confidently.

Two rules fix it permanently:

  • Compute every rate against the last day the data actually covers, not the wall clock.
  • Project every date forward from that same anchor — and put the anchor on the screen where the reader can see it.

A forecast that won't tell you how old its inputs are is asking to be trusted on the one point where it's weakest.

04What actually counts as available

Not every unit attached to an ASIN can serve the next customer. Before any projection, separate:

  • Sellable — the only units that satisfy demand today.
  • Reserved — allocated to orders, in transfer between fulfilment centres, or in processing. Real, but not yet available.
  • Unfulfillable — damaged, expired, stranded. Present in the account, absent from the forecast.
  • Inbound — not stock. Not nothing either. See the next section.

Two of these routinely inflate a forecast. Counting reserved units as sellable buys you a few phantom days; counting inbound as on-hand buys you weeks of them.

05Inbound is a date, not a quantity

The common shortcut — add on-hand and inbound, divide by velocity — produces a number that is almost never right, because it assumes 800 units on a boat can be sold today.

Model stock as a curve instead. It falls at your velocity, steps up on the arrival date, and falls again.

AB-2210 — projected sellable units

At the weighted velocity of 44.3/day, anchored to the last data date.

0 400 800 1,200 1,240 on hand 5 Sep · last data date +800 inbound lands 2 Oct · 44 units left Out of stock ≈ 21 Oct 5 Sep 19 Sep 3 Oct 17 Oct 31 Oct
The shortcut — (1,240 + 800) ÷ 44.3 — gives 46 days of cover. The curve gives the same end date but shows the thing that matters: on 1 October this SKU is down to 44 units. A shipment three days late is a stockout, not a rounding error.

That near-miss is invisible in any calculation that treats inbound as a lump sum. It's the single strongest argument for modelling inventory as a timeline.

And it means a delay notification is a forecast input, not an inconvenience: push the arrival by ten days and the projection should change the moment you record it.

06Lead time is a range you've already measured

Most reorder errors trace back to a lead time somebody typed in once, optimistically, in a different year.

You don't need to estimate it. You've run the orders. Measure from order placed to units sellable — the full chain, not just the factory:

production 20d + prep 5d + transit 25d + receiving 5d

For AB-2210 the last five orders took 48, 53, 59, 67 and 72 days. The average is 60. Reaching for the average alone throws away the most useful thing in that list — the spread.

average lead time  = 60 days
worst observed     = 72 days
lead-time buffer   = 72 − 60 = 12 days

Those twelve days aren't padding. They're the observed cost of your supplier being their usual self on a bad month. Add a demand-variability allowance on top if the SKU is volatile.

Whatever you choose, the assumption belongs on screen. A reorder date computed from a lead time the reader can't see is a number they have no way to disagree with — and the seller usually knows something the system doesn't.

07The whole calculation, on one SKU

Now every input has a defensible value. The reorder point is what you must not fall below:

reorder point = velocity × (lead time + buffer)
              = 44.3 × (60 + 12)
              = 3,190 units

AB-2210 has 1,240 on hand and 800 inbound. Even counting both, it sits at 2,040 — well under 3,190. It crossed its reorder point some time ago.

The date says the same thing:

projected stockout  ≈ 21 Oct   (from the curve above)
minus lead time         60 d
minus buffer            12 d
──────────────────────────────
order-by date       ≈ 10 Aug   — 30 days ago

This is the outcome the article has been building toward, and it's the common one: a SKU that looks healthy — six weeks of cover, a shipment on the water — is a month late on its next order. Days of cover hid it. Only the date exposed it.

When that happens, the questions change. Not "should I order?" but: how much can I expedite, what does air freight cost against the stockout it prevents, and do I split the order to get something moving now.

08Show the range, not the date

A system that prints "Reorder by 10 August" has told the truth and created a false impression at the same time. The number looks like a fact. It's a conclusion resting on a velocity assumption that could reasonably have been three other values.

So show what the assumption is worth:

Conservative 30/day — the 90-day window, if the acceleration is temporary order by 1 Sep
Working 44.3/day — the weighted blend, the current default order by 10 Aug
High demand 47/day — the 7-day window, if the trend holds order by 8 Aug

Read that block for a second and you learn something no single date could tell you: the window debate doesn't matter here. Every plausible assumption puts the order-by date in the past. The velocity question is real, but it isn't this SKU's question — it only decides the quantity now, not whether to act.

On another SKU the same three rows might span six weeks and straddle today. Then the assumption is the decision, and it deserves an afternoon. Showing the spread is what lets a seller tell those two situations apart in a glance.

09What "real-time" should actually mean

Not recalculating every second. Nobody makes a purchase order decision on a five-minute cadence, and a forecast that twitches at every daily fluctuation trains you to ignore it.

Real-time should mean event-driven: the projection updates when something that feeds it changes.

  • New sales data lands
  • Stock moves, or units go unfulfillable
  • An inbound shipment ships, arrives, or slips
  • A purchase order is placed, received, or its dates change
  • You change a lead time or a target cover

And it should be paired with a rule for when to tell you. Alert when a conclusion changes, not when a number does: an order-by date crossing today, a stockout date moving more than a week, a SKU dropping below its reorder point. Everything else is a screen you can visit.

A monthly refresh misses real changes. A constant one manufactures fake ones. The right cadence is a system that recalculates on events and interrupts you on conclusions.

10Keeping score is what makes it improve

Store what you predicted. Compare it to what happened. Without this, a forecasting system never gets better — it just keeps being confidently wrong in the same direction.

Three variances are worth tracking per SKU:

  • Velocity variance — forecast 40/day, actual 52/day. One miss is noise; the same sign four weeks running is a settings problem you can fix.
  • Lead-time variance — every completed PO is a free measurement. Feed it straight back into the average and the worst case.
  • Arrival variance — how often does inbound land on the date you were given? That number is your real buffer requirement, and it's usually larger than the one you chose.

Over time this turns lead time and buffer from opinions into observations. That's the compounding part; the arithmetic was never the hard bit.

11The checklist

  1. Find the last date your sales data covers. Anchor everything to it.
  2. Separate sellable from reserved, unfulfillable and inbound.
  3. Compute the velocity ladder — 7, 30, 60, 90 — and read its shape.
  4. Set the working velocity: weighted by default, overridden only for a cause you can name.
  5. Project stock as a curve, stepping up on each inbound arrival date.
  6. Measure lead time from your own POs; keep the average and the worst case.
  7. Set the buffer from the spread, plus a demand allowance if the SKU is volatile.
  8. Compute the reorder point and the order-by date. Flag anything already past.
  9. Show the order-by date at three velocities, not one.
  10. Recalculate on events; alert only when a conclusion changes.
  11. Record the forecast, and score it later.

Eleven steps, and only two of them are arithmetic. The rest is deciding what you believe and writing it down where you can check it later.

Frequently asked

Which sales velocity window should I use for reorder forecasting?

Default to a weighted blend — 60% of the last 7 days plus 40% of the last 30 — which reacts within a week without being dragged by one unusual day. Override it downward when you can name a temporary cause with an end date, and upward when all four windows show the same direction for two consecutive weeks. Use the faster figure to decide when to act and the slower one to decide how much to buy.

How do I calculate an FBA reorder date?

Project sellable stock forward at your working velocity, stepping up on each inbound arrival date, to find the stockout date. Then subtract the measured lead time and your buffer. If the result is in the past, the order is overdue.

What's the difference between the reorder date and the stockout date?

The stockout date is when you hit zero. The reorder date is when you must act to prevent it — earlier by your entire lead time plus buffer. Sellers who watch only the stockout date consistently order late by the length of their supply chain.

How much safety stock should I carry?

Start from the spread in your own measured lead times: worst observed minus average, expressed in days, times your velocity. Add a demand-variability allowance for volatile SKUs. That gives a buffer grounded in your supplier's actual behaviour rather than a round number.

Should inbound inventory count toward days of cover?

Only from its arrival date. Adding inbound to on-hand and dividing by velocity hides the trough before the shipment lands — which is exactly where a stockout happens if it slips.

Why does my reorder date keep changing?

Because its inputs changed — velocity, an arrival date, a lead time. That's correct behaviour. A reorder date that never moves is one that stopped reading your data.

How often should the forecast update?

On events rather than on a clock: new sales data, a stock change, a shipment shipping or slipping, a PO changing. Alert only when a conclusion changes — a date crossing today, a stockout moving more than a week — not when a number moves.

A deadline, not a date

You can't know that you'll need to reorder an ASIN on a particular day. You can know what today's evidence implies, how sensitive that conclusion is to the one number you're least sure of, and how much time is left before the choice gets made for you.

That's what a reorder forecast is for. Not the date. The deadline, and how much confidence it deserves.

Episode transcript

Read the full transcript

Transcribed automatically from the recording. Speakers are not labelled.

Why can six weeks of cover still mean I'm late?

[00:00]Picture this, you're staring at your operations dashboard. You know, we have all been there. You're looking at one of your top performing SKUs. You've got a container currently on the water, and the screen is giving you this beautiful, comforting green number, like six weeks of inventory cover.

Which feels amazing. It does. You take a breath, and you take a sip of coffee, and you feel like you are totally in control of your supply chain. But, what if I told you that you aren't actually sitting pretty? What if mathematically speaking, you are already a solid month late for your next reorder? And a total stock out is practically guaranteed.

Right. It's a terrifying scenario, but for experienced Amazon sellers, I mean, it is incredibly common. It's the absolute definition of a false sense of security, because the operational mechanics behind that green number, they're often based on fundamentally flawed logic. Exactly. So today's deep dive is for you, the seller who lives and dies by inventory planning. We are breaking down this really fascinating piece of source material from AmazeBase that frankly completely dismantles the way most sellers handle their logistics.

Why is a single reorder date the wrong answer?

It really does. The core argument here is that just asking, "When should I reorder this ASIN?" That is fundamentally the wrong question to begin with. Because a reorder forecast should never ever give you a single falsely precise date, right? Never. It should give you a range, a hard deadline, and absolute transparency regarding the assumptions that actually generated those numbers. I mean, any system spitting out a single date is just hiding its own margin of error.

Yeah. So our mission today is to deconstruct why that single date on your screen is lying to you. Think about it like this. Trusting a single reorder date is kind of like trusting your car's miles to empty gauge while you're driving up a steep mountain. Oh, that's a great analogy.

[01:53]Right. The math inside the car's computer might technically be right based on how you were driving on the highway an hour ago. But the context of the mountain you're currently climbing, that's gonna leave you stranded on the side of the road. Because the terrain has changed, but the gauge hasn't caught up.

What is the difference between the stock-out date and the order-by date?

Exactly. So, to understand why that gauge is broken in the Amazon ecosystem, we have to start by clarifying what we are actually solving for. The AmazeBase article makes this brilliant distinction right out of the gate, separating two fundamentally different timelines. Okay, break that down for us. So most sellers are asking their systems, "When do I run out of stock?" Right? That is the stock out date. It's the exact moment your usable inventory hits zero under some specific demand assumption. But the system needs to be answering a much harder question.

Yes. Identifying the order by date. Right, the order by date. That is the very last day you can safely place a purchase order with your factory and still have those units physically land in a fulfillment center before the stock out actually happens. And what separates those two dates is the entire length of your supply chain, plus whatever safety buffer you've built in. So let's go back to that opening scenario.

Okay. Imagine you have a supplier with a 60 day lead time that covers everything from wiring the deposit to the moment Amazon finally scans the boxes into sellable inventory. If you're looking at an ASIN that currently has 45 days of inventory cover, you're not comfortable. You are already 15 days late. You're in the red. Yeah. So essentially sellers are managing their business to the stock out date when they really should be managing to the order by date. Exactly. I think this is where the mental shift has to happen. Sellers often manage to the stock out date because that is what native platforms naturally highlight. You know, the dashboard turns red when the stock gets low. Right. But waiting for your inventory to look low is a strategy that only works if your overseas manufacturing is somehow faster than your customers clicking buy now.

[03:50]Which obviously is impossible. Completely impossible. The reality of global logistics means your lead time will always dwarf your immediate stock on hand visibility. You basically have to forecast backwards from the anticipated zero hour, not forwards from today's warehouse count. But here's the complication, right? Calculating that order by date requires pulling historical sales data to project your future velocity. And the source material highlights this silent error in how most software and honestly, most sellers using spreadsheets calculate that data.

What is the anchor date trap?

Oh, this is huge. It's an error that ruins the math before you even begin. The AmazeBase piece calls it the anchor date trap. The trap is the clock itself. Tell me more about that. It's such a critical operational insight. It explains why forecasts must be anchored to the last date the sales data actually covers, not today's wall clock. I mean, we all know that API syncs delay.

Oh, constantly. We know Amazon's inventory ledger takes time to settle. Pending orders might sit in a reserve state for days before they officially clear into your shipped revenue metrics. Right. So let's give a concrete example to ground this. The article uses a specific worked example throughout a SKU they call AB 2210. Okay, perfect. Let's assume today is September 17th. But your actual transaction data, because of those settlement delays you just mentioned, only goes up to September 5th. Right. So if your system or your spreadsheet calculates your last 30 days of sales against today's date, September 17th, it takes 18 days of real sales data, and then it averages it with 12 days of empty zero data days. Because the data for those 12 days literally hasn't populated in the system yet. It's just an invisible vacuum. Exactly. The math executes perfectly, but the logic is entirely flawed. The AmazeBase breakdown shows that having just 12 days of lag inside a 30-day window understates your actual sales velocity by roughly 40%.

[05:55]40%? Wow. Yeah. That is a massive margin of error. It's a subtle data issue, but the compounding effect on a spreadsheet must be disastrous. Think about what that does to your dashboard. The system takes your physical stock, divides it by a velocity that is 40% too slow, and gives you this falsely green, incredibly comfortable forecast. You think you have weeks of runway for a SKU that is, in reality, about to go dark. This reminds me, so many software platforms sell themselves heavily on being real time.

You know, you log in and you just assume the number on the screen reflects the reality of the business up to the very minute. But real time software only processes the data it has received. If Amazon hasn't finalized the settlement report, the software is doing real time processing on incomplete information. It's like checking your bank balance on Friday based on checks that cleared last Monday.

That is exactly it. You look at your app, you see a big number, you go out to a nice dinner, but you're actually overdrawn because your mortgage payment hasn't populated yet. Which is why AmazeBase lays down a strict rule here. You have to compute every rate against the last day the data actually covers and project every date forward from that exact same anchor.

Why isn't one 30-day average enough?

Makes sense. And crucially, that anchor date needs to be visible on the screen. If a forecast won't tell you how old its inputs are, it's asking you to trust it on the exact point where it is the weakest. Okay, so even with the anchor date fixed, we hit another wall. Fixing the clock doesn't really help if your speed calculation is broken, because demand is not static. Not at all. Like if you are selling winter coats in October, your velocity today is fundamentally different from your velocity in August. Relying on a single window of time, like just a flat 30 day average, completely flattens out reality. This introduces what the article calls the velocity ladder. It forces the seller to look at the shape of their demand rather than just pulling a single historical average and crossing their fingers. So let's look at the shape of our SKU AB 2210. With the data anchored properly on September 5th, the article pulls four different trailing windows. They look at the 90 day, 60 day, 30 day and 7 windows. And there's a really clear trend.

What does it mean when every shorter window is faster?

[08:06]Let's hear it. The 90-day window is the slowest. The 60-day is a bit faster. The 30-day is faster still, and the 7-day window is the fastest of all, sitting at over 47 units a day. Right, so every shorter window is faster than the one before it. In mathematics, this is called a monotonic ladder.

A monotonic ladder. Yeah. Okay. It basically means the data's moving in one continuous direction without reversing. The signal that shape provides is fascinating. A monotonic ladder suggests you are looking at real sustained acceleration, not just a blip on the radar. Exactly. A temporary spike, say an influencer, mentions your product on TikTok over the weekend. That would throw off your seven day average completely. Right, it would look huge. But your 30, 60, and 90-day averages would remain relatively flat because that volume gets diluted over the longer timeframe.

Which velocity should go into the order?

When the whole ladder moves together in a monotonic sequence, the trend is undeniable. You are experiencing a systemic change in organic demand. But the challenge for the operator is figuring out which of those numbers actually goes into the purchase order math, right? Totally. And the source material suggests a really smart default blend. Relying solely on the 7-day window makes your purchasing way too twitchy.

Yeah. But relying purely on the 30-day window drags you down with historical data that no longer reflects your current momentum. So they recommend taking 60% of the seven-day velocity and adding it to 40% of the 30-day velocity. So that creates a weighted blended velocity. It leans heavily on recent data, so you remain responsive. But it has enough historical gravity that you won't get completely derailed by a single weird Tuesday.

Why use the fast velocity for timing and the slow one for quantity?

Exactly. For AB 2210, that blended calculation gives us roughly 44 units a day. This leads to the golden rule of this entire framework. Oh, I love this part. You use the faster velocity assumption to decide when to act, and you use the slower velocity assumption to decide how much to buy. The risk asymmetry there is brilliant. We should definitely explore the operational consequences of separating those two decisions. Because the risks are not symmetrical at all. If you act early based on a fast velocity, what's the worst outcome?

[10:15]You incur a few extra days of warehouse carrying costs, Amazon's storage fees might tick up slightly. But if you buy too much inventory based on a fast velocity that turns out to be a temporary spike, you trap tens of thousands of dollars in cash. You might not see that capital again for six months and you risk punitive long-term storage fees. But operators face situations constantly where they know the data is skewed, right? Like what if that seven day spike is just the tail end of prime day or a specific coupon you ran?

When should I ignore a short-term spike?

The velocity ladder might look steep, but you absolutely know it isn't real long-term demand. In those specific scenarios, you override the default blend. The rule laid out is to discount the short window when you can name the cause. And that cause has a definitive end date.

Like a holiday. Exactly. A coupon, a competitor temporarily out of stock, a seasonal holiday, these are discrete events, not baseline demand. You manually reprice your forecast to your pre-event baseline. Got it. However, if the ladder is monotonic, the gap has persisted for two consecutive weeks and you have no idea why you trust it. That usually indicates a permanent shift in your algorithmic rank. Okay, so we have our properly anchored data and we have a highly accurate blended speed of sales. But the speed is only half the equation, isn't it?

Why shouldn't I count inbound stock as stock on hand?

Yes, it is. We have to look at what we actually have available to sell our inventory levels. And here's where operational discipline often just completely falls apart, Because sellers constantly hallucinate their stock by treating boats as warehouses. The inbound mirage. It is a massive blind spot. To get this right, you have to explicitly decouple your inbound shipments from your sellable units on hand. You've got four types of inventory. Sellable, reserved, unfulfillable, and inbound. And unfulfillable we just ignore entirely, right?

[12:07]Right, ignore it. But treating inbound inventory as on hand stock is disastrous. Let's run through the shortcut formula that almost everyone uses. You take your on hand sellable stock, you add your inbound stock, and you divide by your daily velocity. So for our SKU AB2210, we have 1,240 units on hand and we have 800 units inbound on a boat. Okay. The shortcut logic says we have 2,040 units total divided by 44 units a day giving us a very comfortable 46 days of inventory cover. But that 46 days of cover is almost entirely fiction. How so? Well, it assumes that those 800 units sitting in a shipping container in the middle of the Pacific Ocean can be picked, packed and shipped to a customer in Ohio this afternoon. Right, which they obviously can't. You cannot model inbound inventory as a lump sum added to your current stock. You have to model your physical inventory as a curve dropping over time with a sudden vertical step up on the exact anticipated arrival date.

Let's trace that conceptual curve for AB 2210. Our data is anchored on September 5th with 1,240 physical units. We are burning through 44 units a day. Our inbound shipment of 800 units is expected to arrive and become sellable on October 2. If you project that daily burn rate forward, something absolutely terrifying happens on the curve. On October 1, the day before that boat is supposed to arrive, your sellable stock drops to exactly 44 units. You are literally one day's worth of sales away from absolute zero. So the dashboard told you that you had 46 days of cover.

But the curve tells you the operational truth. If that inbound shipment slips by just three days, you are going into a complete stock out. It is the sheer panic of realizing you have over 2,000 units legally attached to your company, but you are functionally out of stock. And a three-day slip is nothing in global logistics. A customs delay, a port strike, a missed warehouse appointment at an Amazon fulfillment center, any of those triggers a stock out. Inbound inventory isn't a quantity. It's a date. Inbound is a date. That mantra needs to be at the center of every operations manager's process.

[14:25]It means that an email from your freight forwarder saying your shipment is delayed isn't just an administrative annoyance, you know? Right, changes everything. It is a critical forecast input. If that arrival date pushes by 10 days, your projection needs to instantly update because you are now forecasting a seven-day stock out and a total loss of organic momentum. So, because we know that inbound shipments will eventually slip, we have to protect ourselves with a buffer. But sellers usually just guess this number, right? They type 14 days into their software because it feels like a nice safe cushion.

How big should my safety buffer be?

Which is so dangerous. The AmazeBase piece shows a much more rigorous way to do this, building a buffer from reality, using your own historical purchase orders. Moving from optimistic guesses to empirical measurement is key here. Most reorder errors trace back to a lead time that somebody typed into a backend system two years ago during a totally different economic climate.

Exactly. Factory power outages happen. Raw material shortages occur. Ocean freight carriers roll cargo. You don't need to guess your lead time or your buffer. You just need to look at your own operational data. And they stress measuring the full chain. Order placed to units sellable, not just the factory production time. It's production plus prep, plus ocean transit, plus Amazon's notoriously unpredictable internal receiving times. Let's examine the last five purchase orders for AB 2210. The actual elapsed times were 48 days, 53 days, 59 days, 67 days, and 72 days. If you take the average of those five, you get 60 days. Which is the exact number most off the shelf software automatically calculates and uses as your static lead time.

But relying solely on that average completely obscures the volatility of your supply chain, doesn't it? Completely. The most valuable piece of data in that list is the spread between the average, which is 60, and the worst observed case, which is 72. The spread is 12 days.

What is my reorder point and my order-by date?

[16:24]That becomes your buffer. Yes. Those 12 days aren't arbitrary padding. They are the mathematically measured cost of your supplier being themselves on a bad month. It is the quantifiable reality of your specific logistics chain over the last year. So we can bring this all together and calculate the final reality for AB 2210. The reorder point is our velocity, roughly 44 units a day, multiplied by our lead time plus our buffer. So 60 days plus 12 days is 72 days. 44 units times 72 days is roughly 3,190 units.

Meaning the absolute minute your total inventory, both on hand and inbound, drops below 3,190 units, you must place your next order. But wait, our total inventory for AB 2210 is only 2,040 units. We have 1,240 on hand and 800 on the boat. We are over 1,000 units below the reorder point. The SKU crossed its reorder point long ago. If you project the stockout date using our curve and subtract the 60-day lead time and the 12-day buffer, you find the real order by date. For this SKU, the order by date was August 10th. Almost a month in the past. Yeah. If a seller runs this math and suddenly realizes they are a month late, The standard planning phase is completely over. What do they actually do right now?

What do I do if I'm already late?

Well, getting this right is vital because it shifts your entire operational focus. You stop asking, "Should I order?" and you immediately transition into crisis management. You have to ask, "How much of this new order needs to go via air freight?" Right. Checking costs. What does air freight cost compared to the lost organic rank of a three-week stock out? Can my factory split the order and send 500 units express right now? You move from forecasting into immediate triage. The worst thing an operator can do is place a standard ocean freight order today, blindly trusting a green dashboard, only to watch the listing go dark in three weeks and stay dark for a month and a half.

Why should a forecast show a range instead of one date?

[18:25]Wow. As we wrap this up, the AmazeBase piece ends with a really strong argument about how software should display this information back to the seller. It shouldn't just be one big bold date on a screen. It needs to be a range. A single date creates a false impression of certainty. It is a conclusion resting on assumptions about demand that might be completely wrong. The dashboard needs to show the seller what those assumptions are actually worth in real time. Right, and they show it across the different velocity windows we discussed earlier.

The conservative assumption, which was the 90-day window, gives an order by date of September 1st. The working assumption, or blended 44 units a day, gives an order by date of August 10th. And the high demand assumption, the 7-day spike, gives an order by date of August 8th. Seeing all three rosette ones is incredibly clarifying.

Definitely. Because by looking at that spread, you learn something no single date could ever communicate. The debate over which velocity window to use doesn't even matter for this SKU right now. Because all three scenarios are in the past. Whether you are conservative or aggressive in your demand assumptions, the conclusion is identical. You are incredibly late. Call the factory. Exactly. The velocity question only dictates the quantity of the PO now, not the timing. Now if those three dates span six weeks and straddled today's date, then the assumption becomes the core of the decision.

What should a real-time dashboard actually do?

Right. You would need to spend an afternoon digging into the data to figure out which velocity is real before tying up capital. Showing the range allows a seller to tell those two completely different situations apart in a single glance. It really redefines what a, quote, "Real-time dashboard dashboard should actually be. A twitchy, minute by minute clock updating every time you sell a single unit just causes alert fatigue. It trains the operator to ignore the noise. Real time should mean event driven. The projection should only update when a variable that feeds it fundamentally changes.

[20:23]A shipment slips by a week at the port. A purchase order is delayed by the factory. A whole new batch of sales data settles from the API. So the system should only alert you when a conclusion changes, like a stock out date moving significantly, or a SKU dropping past its reorder point. Alert when a conclusion changes, not when a number moves. It keeps the focus entirely on action rather than just raw data consumption. It respects the operator's time and focuses their attention on actual risk. Bringing it all back. If a beautifully precise green reorder date on a dashboard was actually hiding a month of supply chain delays, it really makes you wonder. What other perfectly precise metrics on your operational dashboard are giving you a false sense of security right now.

What other comforting numbers could be misleading me?

A great question. Are your blended profit margins hiding unsellable return costs? Are your advertising metrics hiding algorithmic cannibalization? It might be time to stop trusting the gauge, get out of the car, and look closely at the assumptions beneath the exactness, especially when you're driving up that mountain.

Question everything. Especially the metrics that look the most comforting.