The Inventory Formula Amazon's Top 1% Sellers Use (That Nobody Talks About).

There are two kinds of Amazon sellers. The kind who reorders when stock "feels low." And the kind who reorders at exactly 847 units because they calculated that number using four variables updated every week.

The first kind runs out of stock 6-8 times a year. Loses organic ranking each time. Spends weeks recovering sales velocity after each stockout. Bleeds money.

The second kind runs out of stock 0-1 times per year. And when they do, it is because something genuinely unpredictable happened, not because their math was wrong.

The difference is one formula. Here it is.

The Formula

Most inventory guides give you this:

Reorder Point = (Average Daily Sales x Lead Time) + Safety Stock

That is the kindergarten version. It works if you sell the same amount every day, on one channel, with a supplier who never misses a delivery window. In other words, it works in a textbook. Not in reality.

Here is what top sellers actually use:

Reorder Point = (Weighted Channel Velocity x Adjusted Lead Time x Seasonal Coefficient) + Dynamic Safety Stock

Four components. Each one matters. Let me break them down.

Component 1: Weighted Channel Velocity

If you sell on multiple channels, "average daily sales" is a blended number that hides important information. Amazon might sell 30 units/day with wild swings. Your Shopify store might sell 15 units/day with steady consistency. eBay might sell 5 units/day with spikes every weekend.

Blending these into "50 units/day" ignores the fact that each channel has different volatility. And volatility is what kills you, not averages.

How to calculate it:

1. Track daily unit sales per channel for the last 30 days
2. Calculate the average daily sales per channel
3. Sum them for your total weighted daily velocity

Channel	Avg Daily Sales	Std Deviation	Coefficient of Variation
Amazon	30	12	0.40
Shopify	15	4	0.27
eBay	5	3	0.60
Total	50	,	,

Notice eBay's coefficient of variation (standard deviation / average) is 0.60. That is highly volatile relative to its volume. Amazon is at 0.40. Shopify is the most predictable at 0.27. This variance data feeds directly into your safety stock calculation.

If you use a tool like Nventory that tracks sales velocity per channel, this data is already being collected. You just need to extract it for the formula.

Component 2: Adjusted Lead Time

Lead time is not the number your supplier told you. It is the number you have actually experienced.

Your supplier says 14 days. But over the last 6 orders, your actual lead times were: 14, 16, 14, 21, 15, 17 days. Your average lead time is 16.2 days. Your maximum was 21 days. Big difference from the quoted 14.

How to calculate it:

1. Record actual lead times for your last 6-10 orders from each supplier
2. Calculate the average: this is your Adjusted Lead Time for the formula
3. Calculate the standard deviation: this feeds into safety stock
4. Note the maximum: this is your worst-case planning scenario

Order #	Quoted Lead Time	Actual Lead Time	Variance
1	14 days	14 days	0
2	14 days	16 days	+2
3	14 days	14 days	0
4	14 days	21 days	+7
5	14 days	15 days	+1
6	14 days	17 days	+3

Average lead time: 16.2 days. Standard deviation: 2.6 days. If you plan using 14 days, you are understocking every time the supplier is late. And suppliers are late more often than they are early.

Component 3: Seasonal Coefficient

This is the variable that separates amateurs from professionals. Without it, you stock the same amount in July as you do in November. That means dead capital in slow months and stockouts in peak months.

How to calculate it:

1. Pull monthly unit sales for the last 12-24 months
2. Calculate the overall monthly average
3. Divide each month's actual sales by the monthly average

Month	Units Sold	Monthly Average	Seasonal Coefficient
January	980	1,500	0.65
February	1,050	1,500	0.70
March	1,200	1,500	0.80
April	1,350	1,500	0.90
May	1,500	1,500	1.00
June	1,350	1,500	0.90
July	1,200	1,500	0.80
August	1,500	1,500	1.00
September	1,650	1,500	1.10
October	1,800	1,500	1.20
November	2,550	1,500	1.70
December	2,870	1,500	1.91

In December (coefficient 1.91), your reorder point is nearly double what it would be in January (coefficient 0.65). That is a massive difference. Sellers who ignore seasonality either run out of stock in Q4 (losing their biggest sales month) or sit on excess inventory in Q1 (tying up capital for months).

Component 4: Dynamic Safety Stock

Safety stock is your buffer against uncertainty, both demand uncertainty and supply uncertainty. Most sellers set it as a flat number ("keep 50 extra"). That is lazy math that either costs you money or costs you sales.

The formula:

Safety Stock = Z x √(Lead Time x σ²demand + Average Daily Sales² x σ²lead time)

Where:

• Z = Z-score for your target service level (1.65 for 95%, 2.05 for 98%, 2.33 for 99%)
• σ demand = standard deviation of daily demand (combined across channels)
• σ lead time = standard deviation of lead time in days

This accounts for both demand variability AND supply variability. Most simple safety stock formulas only account for one.

Worked Example: 50 Units/Day, 14-Day Quoted Lead Time

Let me run through the complete formula for a real product.

Given:

• Total daily sales: 50 units/day (across Amazon, Shopify, eBay)
• Standard deviation of daily demand: 13 units
• Quoted lead time: 14 days
• Actual average lead time: 16.2 days
• Standard deviation of lead time: 2.6 days
• Current month: October (seasonal coefficient: 1.20)
• Target service level: 95% (Z = 1.65)

Step 1: Calculate Safety Stock

Safety Stock = Z x √(Lead Time x σ²demand + Daily Sales² x σ²lead time)

= 1.65 x √(16.2 x 169 + 2,500 x 6.76)

= 1.65 x √(2,737.8 + 16,900)

= 1.65 x √19,637.8

= 1.65 x 140.1

= 231 units

Step 2: Calculate Lead Time Demand

Lead Time Demand = Average Daily Sales x Adjusted Lead Time

= 50 x 16.2

= 810 units

Step 3: Apply Seasonal Coefficient

Seasonally Adjusted Lead Time Demand = 810 x 1.20

= 972 units

Step 4: Calculate Reorder Point

Reorder Point = Seasonally Adjusted Lead Time Demand + Safety Stock

= 972 + 231

= 1,203 units

When your inventory hits 1,203 units, place the order. Not 500. Not "when it feels low." 1,203.

Compare This to the Simple Formula

If you used the basic formula (Average Daily Sales x Quoted Lead Time + 50 safety stock):

= (50 x 14) + 50 = 750 units

That is a 453 unit difference. The simple formula would have you reordering so late that you would run out of stock before the shipment arrived, especially in a peak month when demand is 20% above average and the supplier delivers 2 days late.

How the Formula Changes at Different Scales

Daily Volume	Simple Reorder Point	Dynamic Reorder Point	Difference
10 units/day	190	287	+97 units
25 units/day	400	628	+228 units
50 units/day	750	1,203	+453 units
100 units/day	1,450	2,341	+891 units
200 units/day	2,850	4,567	+1,717 units

The gap widens as volume increases. At 200 units/day, the simple formula has you reordering 1,717 units too late. That is 8.5 days of stock you do not have. Eight and a half days of stockout on your biggest products.

The Economic Order Quantity Connection

Knowing when to reorder is half the equation. Knowing how much to reorder is the other half.

Economic Order Quantity = √(2 x Annual Demand x Order Cost / Holding Cost per Unit)

For our 50-unit/day product:

• Annual demand: 18,250 units
• Order cost (shipping, inspection, processing): $350 per order
• Annual holding cost per unit: $2.40 (storage + insurance + capital cost)

EOQ = √(2 x 18,250 x 350 / 2.40) = √5,322,917 = 2,307 units per order

So you reorder 2,307 units when inventory hits 1,203. This gives you roughly 46 days of stock at average velocity, which means you are placing an order roughly every 6-7 weeks.

Why Static Minimums Are Killing You

I talk to sellers every week who say: "I reorder when I hit 200 units." Why 200? "That is what I have always done."

That fixed number was probably reasonable at some point, when they sold 15 units/day and their supplier delivered in 10 days. But now they sell 35 units/day (growth) and their supplier takes 18 days (supply chain disruptions). Their reorder point should be 900+. They are still using 200.

Result: they stock out 4-5 times a year. Each stockout on Amazon costs them organic ranking. Each ranking drop costs 2-3 weeks of recovery time. Each recovery period has lower sales velocity, which means lower contribution to next quarter's seasonal forecast, which means they understock again next year. It is a compounding problem.

The top 1% recalculate monthly. They update their lead time data with every supplier order. They adjust seasonal coefficients each year based on actuals. They track channel velocity independently. And they never, ever, use a static minimum.

Implementing This Without a PhD

You do not need a supply chain degree to use this formula. You need a spreadsheet and 30 minutes per month.

Step 1: Build the Data Habit

Track three numbers for every SKU, every month: total units sold per channel, actual lead time per supplier order, and current stock level. That is it. Everything else is derived from these three inputs.

Step 2: Start With Your Top 10 SKUs

Do not try to apply this to 500 SKUs on day one. Your top 10 products by revenue probably represent 40-60% of total sales. Get those reorder points right and you have solved the majority of the problem.

Step 3: Set Calendar Reminders

First Monday of every month: update the spreadsheet, recalculate reorder points, place any orders that are due. This takes 30-45 minutes once you have the template built. For your top 3 SKUs, check weekly.

Step 4: Automate When Ready

Once you trust the formula, look for tools that can automate the inputs. If your OMS already tracks sales velocity per channel and provides low-stock alerts, tools like Nventory do this natively, you can feed that data directly into your reorder calculations instead of pulling it manually.

The formula is not complicated. It is just rigorous. The difference between the top 1% and everyone else is not intelligence or resources. It is the willingness to treat inventory math as a weekly discipline rather than an annual guess.

Your competitors who never stock out? They are not lucky. They are doing this math. Now you can too.