Most e-commerce sellers order reactively. They notice inventory is getting low, or a team member mentions an item is almost gone, and they place an order. Sometimes they catch it in time. Sometimes they don't. The item goes out of stock, sales stop, and they scramble to reorder and wait on restocking.
The other version of this problem is the over-correction: they get burned by a stockout, so they over-order next time and end up sitting on six months of inventory that stopped moving.
A reorder point system doesn't solve every inventory problem, but it solves this one. It takes ordering out of reactive mode and gives you a clear trigger: when this item hits this quantity, you order.
What a reorder point actually is
A reorder point is a quantity threshold, not a date or a gut feeling. When your on-hand count for an item drops to the reorder point, you know it's time to place an order. Before that, you don't need to. After that, you're already behind.
That's the whole concept. The rest is just getting the numbers right.
The formula
Reorder Point = (Average Daily Sales × Lead Time) + Safety Stock
Average Daily Sales. Take the last 90 days of sales for a given SKU and divide by 90. Ninety days is enough data to smooth out noise while still reflecting recent velocity. If an item launched recently or went through a major promotion in that window, account for that before using the raw average.
Lead Time.This is the full cycle — from when you place the order to when units are physically on your shelf and ready to ship. Include transit, receiving, and any processing delays. If your supplier quotes 7 days but it's typically 10, use 10. Use the realistic number, not the optimistic one.
Safety Stock.A buffer for when things don't go as planned — a late shipment, a spike in demand, a receiving day that gets pushed. For fast-moving items, 10 to 14 days of sales is a reasonable buffer. For slow movers, zero is fine. The goal is protection against your most likely disruptions, not against every possible scenario.
A worked example
Say a product sells an average of 8 units per day based on the last 90 days. Your supplier takes 12 days door-to-door. You want a 10-day safety stock buffer for this item.
(8 × 12) + (8 × 10) = 96 + 80 = 176 units
When your count hits 176, you order. Not when it feels low. Not when someone mentions it. At 176.
Building it in a spreadsheet
You don't need software for this. A spreadsheet with these columns is enough to manage reorder points for most sellers at 100–500 active SKUs:
| Column | What it is |
|---|---|
| SKU / Item | Your product identifier |
| Avg daily sales | Total units sold (last 90 days) ÷ 90 |
| Lead time (days) | Realistic door-to-door from your supplier |
| Safety stock (days) | Buffer days you want for this item |
| Calculated ROP | (Avg daily × lead time) + (avg daily × safety days) |
| Current qty | Your actual on-hand count |
| Order flag | Flag if current qty is at or below the ROP |
The order flag column is the payoff. A formula that flags any item where current quantity is at or below the reorder point gives you something you can filter on. Once a week, filter to show anything flagged, review each item, and place orders as needed.
What to watch out for
Velocity changes.A 90-day average works until your sales pattern shifts. An item selling 5 units a day three months ago might be selling 15 now — or 2. Update your averages when velocity changes meaningfully. This isn't a system you set and forget.
Seasonality.If you sell items with strong seasonal patterns, the reorder point needs to reflect that. A summer product calculated during winter will have a reorder point that's too low. Recalculate more frequently during transition periods or build in a seasonal adjustment factor.
On-hand accuracy. The system only works if the quantity column is accurate. If your counts are off — receiving not entered, returns unprocessed, picks not decremented — the reorder point will flag at the wrong time. Fix inventory accuracy first. A reorder point system built on bad counts doesn't help you; it just automates the wrong decisions faster.
Don't automate the ordering yet
The system flags what to look at. You still supply the judgment.
Before placing an order on a flagged item, check: Is there anything unusual about recent sales velocity? Is a recent order already in transit that will cover this? Is the supplier reliable right now? Is this an item where the average was skewed by a promotion?
Manual review also keeps you familiar with your inventory. That familiarity is worth something — you'll catch problems the spreadsheet can't. When you've run the system long enough to trust the numbers and your supplier relationships are stable, that's when you think about automating order triggers.
Start with the spreadsheet. Run it for a few months. You'll know when you need more.