How to Know Exactly When to Reorder Stock (The Formula That Replaced Our Gut Feeling).

Last February, our top-selling product stocked out for 11 days. It sold 42 units a day at a $34 margin. That is $15,708 in lost profit. But the real cost was worse: we lost our organic ranking on Amazon, which took 3 weeks to recover, costing us roughly another $31,000 in reduced sales velocity during the recovery window.

Total cost of one stockout: approximately $47,000.

Why did it happen? Because our reorder process was "Tom thinks we should probably order more soon." Tom was on vacation. Nobody else was thinking about it. The stock ran out on a Thursday. The supplier lead time was 16 days. By the time we noticed and placed an emergency order (at a 22% premium for expedited shipping), we had already lost 11 selling days.

That was the last time we used gut feeling. Now we use a formula. Here is exactly how it works.

The Core Formula

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

When your inventory count drops to this number, you place a purchase order. Not before (tying up cash). Not after (risking a stockout). At this exact number.

The formula has three variables. Each one is simple to understand and surprisingly tricky to calculate correctly. Let me walk through each one.

Variable 1: Average Daily Sales

This seems obvious. Total units sold divided by number of days. But here are the three mistakes almost everyone makes:

Mistake 1: Using Too Short a Window

If you calculate daily sales from last week, you are optimizing for last week's demand. If last week included a promotion, a viral TikTok, or a holiday bump, your "average" is inflated. If last week was slow due to weather or a platform glitch, it is deflated.

The fix: Use 30 days for fast-moving products (10+ units/day) and 60-90 days for slower products. This smooths out short-term noise while still reflecting recent trends.

Mistake 2: Ignoring Stockout Days

If you were out of stock for 5 days last month, those 5 days had zero sales, not because demand was zero, but because you had nothing to sell. Including zero-sale days in your average understates demand.

The fix: Exclude days when you were out of stock or had restricted inventory. If you sold 600 units in 25 selling days (not 30), your daily average is 24, not 20. That 20% difference can mean reordering hundreds of units too late.

Mistake 3: Not Aggregating Across Channels

If you sell on Amazon, Shopify, and eBay, your reorder point needs to cover demand from all three channels combined, because all three channels draw from the same inventory pool.

Channel	Units Sold (30 days)	Daily Average
Amazon	720	24.0
Shopify	330	11.0
eBay	150	5.0
Total	1,200	40.0

If you calculate reorder points using only Amazon sales (24/day), you will stock out 16 units/day faster than expected because Shopify and eBay are also consuming inventory. This is the most common reorder calculation error for multichannel sellers.

If you use a tool like Nventory that aggregates sales across channels, this number is calculated for you. If you are doing it manually, pull sales reports from every channel and sum them.

Variable 2: Lead Time

Lead time is the number of days between placing a purchase order and having sellable inventory on your shelves (or in FBA). It is not just manufacturing or shipping time. It is the full cycle:

Lead Time Component	Typical Duration
Order processing by supplier	1-3 days
Manufacturing or picking	3-14 days
Shipping (domestic)	3-7 days
Shipping (international sea freight)	25-45 days
Customs clearance (if international)	2-5 days
Receiving and quality inspection	1-3 days
FBA inbound processing (if applicable)	3-14 days
Total (domestic supplier, self-fulfilled)	8-20 days
Total (international supplier, FBA)	40-80 days

The range on that international-to-FBA number is enormous. 40 days when everything goes right. 80 days when your container gets delayed, customs flags your shipment for inspection, and Amazon's receiving center is backed up during peak season.

Critical rule: Use your actual average lead time from past orders, not your supplier's quoted lead time. Track every order: date you placed PO, date inventory was sellable. After 5-6 orders, you will have an accurate average.

Variable 3: Safety Stock

Safety stock is your insurance policy. It covers two risks:

1. Demand risk: Customers buy more than your average predicts
2. Supply risk: Your supplier delivers later than your average lead time

The Simple Method

Safety Stock = (Maximum Daily Sales x Maximum Lead Time) - (Average Daily Sales x Average Lead Time)

For our product:

• Maximum daily sales (busiest day in 90 days): 62 units
• Maximum lead time (longest order): 22 days
• Average daily sales: 40 units
• Average lead time: 16 days

Safety Stock = (62 x 22) - (40 x 16) = 1,364 - 640 = 724 units

That is a lot of safety stock. This method is conservative, it protects against the absolute worst case on both demand and supply simultaneously. Good for critical products where a stockout is catastrophic. Expensive for products where occasional stockouts are acceptable.

The Statistical Method (More Precise)

If you want to dial in the right balance between protection and capital efficiency, use the service level approach:

Safety Stock = Z x σ daily demand x √Lead Time

Where Z is the Z-score for your target service level:

Service Level	Z-Score	What It Means
90%	1.28	In stock 90% of the time. Acceptable for long-tail SKUs.
95%	1.65	In stock 95% of the time. Standard for most products.
98%	2.05	In stock 98% of the time. For high-margin or critical products.
99%	2.33	In stock 99% of the time. Only for products where stockout is unacceptable.

For our product:

• Z (95% service level) = 1.65
• Standard deviation of daily sales = 11 units
• Average lead time = 16 days

Safety Stock = 1.65 x 11 x √16 = 1.65 x 11 x 4 = 72.6 ≈ 73 units

Notice the difference: 724 units (simple method) vs. 73 units (statistical method at 95% service level). The simple method is 10x more conservative. For a product that costs $8/unit, the simple method ties up an extra $5,208 in safety stock capital. Whether that is worth it depends on what a stockout costs you.

Putting It All Together: The Complete Calculation

Let me run through the full formula for our product:

Inputs:

• Average daily sales (all channels): 40 units
• Average lead time: 16 days
• Safety stock (95% service level): 73 units

Reorder Point:

= (40 x 16) + 73

= 640 + 73

= 713 units

When inventory hits 713 units, we order. The 640 covers expected demand during lead time. The 73 covers unexpected spikes. If our supplier delivers on time and demand stays average, we will have 73 units left when the new stock arrives. If demand spikes 20% during lead time or the supplier is 2 days late, we still have buffer.

Adding Seasonal Adjustment

The formula above assumes demand is constant year-round. It is not. Here is how to adjust:

Month	Historical Sales	Monthly Average	Seasonal Coefficient	Adjusted Daily Sales	Adjusted Reorder Point
January	780	1,200	0.65	26	489
April	1,080	1,200	0.90	36	649
July	1,200	1,200	1.00	40	713
October	1,560	1,200	1.30	52	905
November	2,040	1,200	1.70	68	1,161
December	2,280	1,200	1.90	76	1,289

Look at the reorder point range: 489 units in January vs. 1,289 in December. If you use the flat 713 all year, you overstock by 224 units in January (tying up $1,792 in unnecessary inventory) and understock by 576 units in December (guaranteeing a stockout during your biggest sales month).

Seasonal adjustment is not optional. It is the difference between running out of stock on Black Friday and cruising through it.

Adjusting for Supplier Reliability

Some suppliers are like clockwork. Others treat delivery dates as suggestions. You need to factor their reliability into your lead time:

Supplier Type	Quoted Lead Time	Actual Range	Use in Formula
Highly reliable (on time 90%+)	14 days	13-15 days	15 days (average)
Moderately reliable (on time 70-90%)	14 days	12-20 days	17 days (average + buffer)
Unreliable (on time under 70%)	14 days	14-30 days	22 days (75th percentile)

For unreliable suppliers, use the 75th percentile of actual lead times, not the average. This means 75% of orders arrive before you run out. For the 25% that are later, safety stock covers the gap.

A Real-World Calculator Walkthrough

Let me calculate reorder points for three different products at different scales:

Product A: Low volume, domestic supplier

• Daily sales: 8 units/day (Shopify only)
• Lead time: 7 days
• Std dev of daily demand: 3 units
• Service level: 95% (Z = 1.65)

Safety Stock = 1.65 x 3 x √7 = 1.65 x 3 x 2.65 = 13 units

Reorder Point = (8 x 7) + 13 = 56 + 13 = 69 units

Product B: Medium volume, multichannel

• Daily sales: 40 units/day (Amazon 24, Shopify 11, eBay 5)
• Lead time: 16 days
• Std dev of daily demand: 11 units
• Service level: 95% (Z = 1.65)

Safety Stock = 1.65 x 11 x √16 = 73 units

Reorder Point = (40 x 16) + 73 = 640 + 73 = 713 units

Product C: High volume, international supplier, FBA

• Daily sales: 120 units/day (Amazon 80, Shopify 25, Walmart 15)
• Lead time: 55 days (international + FBA inbound)
• Std dev of daily demand: 28 units
• Service level: 98% (Z = 2.05, high-margin product)

Safety Stock = 2.05 x 28 x √55 = 2.05 x 28 x 7.42 = 426 units

Reorder Point = (120 x 55) + 426 = 6,600 + 426 = 7,026 units

Product C has a reorder point of 7,026 units. At $6/unit cost, that is $42,156 in inventory you need on hand just to trigger a reorder. This is why international sourcing for high-volume products demands serious working capital planning.

When the Formula Saved Us $47,000

Remember that stockout I mentioned at the top? Here is what happened after we implemented the formula.

Three months later, we noticed our best product's inventory hit the reorder point, 713 units, on a Tuesday morning. Under the old system, nobody would have noticed for another week. We placed the PO immediately.

The supplier ended up being 3 days late. During those 3 extra days, we sold 132 units (demand had spiked slightly above average). When the new stock arrived, we had 41 units left, well within our safety stock buffer.

Zero stockout. Zero lost ranking. Zero emergency shipping premium. The formula earned back its development cost in one cycle.

Implementation: Start Here

You do not need software to start. A spreadsheet works. Here is what to build:

1. Column A: SKU name
2. Column B: Average daily sales (last 30-90 days, across all channels)
3. Column C: Average lead time (from actual orders, not supplier quotes)
4. Column D: Standard deviation of daily sales
5. Column E: Safety stock = 1.65 x D x √C
6. Column F: Reorder point = (B x C) + E
7. Column G: Current stock level
8. Column H: Status (Green if G > F, Yellow if G < F x 1.2, Red if G < F)

Update this weekly. It takes 20 minutes once the template is built. When Column H turns red, order. When it turns yellow, start preparing the PO.

That is it. No gut feeling. No "I think we are running low." Just math that works every time, because $47,000 mistakes only have to happen once before you stop guessing and start calculating.