Every brand eventually asks the same question: "At what stock level should I reorder?" Answer it with a feeling and you'll be early sometimes, late sometimes, and stressed always. Answer it with a reorder point calculation and it becomes a number your whole team can act on without a meeting. This guide covers the reorder point formula, two fully worked examples from real D2C scenarios, and the three mistakes that quietly break the math.
Key takeaways
- Reorder Point = (Daily Sales Velocity × Lead Time) + Safety Stock. Three inputs, one number.
- Velocity must exclude stockout days — calendar averages that include them make every reorder point too low.
- Use real observed lead times, not what the supplier quoted at onboarding.
- Set reorder points per variant, not per product — your Medium sells 3× faster than your XL.
- Recalculate at least monthly; velocity drifts, and a stale reorder point protects a business that no longer exists.
The reorder point formula
A reorder point (ROP) is the inventory level that triggers your next order. It answers one question: how many units will I sell between placing an order and receiving it — plus a buffer for when reality misbehaves?
The three inputs:
- Daily sales velocity — units sold per day, averaged over a recent window (30-90 days), counting only days the product was actually in stock.
- Lead time — days from placing the order to sellable stock in your warehouse. For brands that manufacture, that's the full production cycle: materials, production, QC, transit.
- Safety stock — buffer units for demand spikes and late deliveries. Size it properly with the safety stock formula; don't just guess a round number.
When on-hand stock touches the ROP, you order. Not when it feels low, not at the next monthly review — at the number.
A note on lead time, because it's the input brands most often understate: the clock runs from "we should reorder" to "sellable units on the shelf." For a manufacturer that includes procuring materials (your fabric supplier has a lead time too), the production run itself, QC, inbound transit — and the two days a PO sits in someone's inbox waiting for approval. If your bill of materials isn't mapped, the material half of that clock is invisible, and every ROP you set will be quietly optimistic.
Example 1: A kurta brand with steady demand
A Jaipur apparel brand sells its Classic White Kurta at a steady clip. The numbers:
- Daily velocity: 4.2 units/day (90-day average, stockout days excluded)
- Lead time: 14 days (fabric is in stock with the supplier; cutting, stitching, QC, and delivery take two weeks)
- Safety stock: 21 units (about 5 days of cover for late fabric or a good weekend)
The moment stock hits 80 units, the brand starts its next production run. During the 14-day wait it expects to sell ~59 units, landing at ~21 units — the safety stock — just as fresh stock arrives. If everything goes perfectly, safety stock is never touched. That's fine. It's insurance, not waste.
One wrinkle: this kurta comes in five sizes, and 4.2/day is the product total — Medium alone does 1.7/day while XL does 0.4. The production run the ROP triggers should be split by variant velocity, or it arrives pre-imbalanced: Medium sells out in weeks while XL becomes next season's markdown. More on this under mistake 3 below.
Example 2: A beauty brand riding a viral spike
Now the hard case. A Mumbai beauty brand's vitamin C serum has been doing a comfortable 12 units/day. Then a creator's reel takes off and the last 10 days averaged 19/day. Its contract manufacturer quotes 15 days but has delivered in up to 18 during busy months.
Steady-state math would say: ROP = (12 × 15) + 40 = 220 units. But planning on old velocity during a spike is how viral moments turn into stockouts. The brand recalculates using a short window that reflects the spike, and buffers with worst-case inputs:
The reorder point nearly doubled — from 220 to 432 — because both demand and its uncertainty jumped. Two practical notes: during a spike, recalculate weekly, since velocity is moving; and when the spike fades, let the ROP come back down, or you'll overstock on the way out. This is exactly the situation where a system that recalculates daily earns its keep — Honey Shelf's velocity engine reruns this math every day per variant and moves the countdown for you.
Three mistakes that break reorder points
1. Using calendar averages that include stockout days
The most common and most expensive mistake. Say your serum sells 10/day when available, but was out of stock for a third of last month. Your spreadsheet divides 200 units by 30 days and reports 6.7/day. Your true velocity is 10/day. Every ROP built on 6.7 fires 33% late, causing the next stockout, which drags the average down again — a doom loop where each stockout makes the next one more likely. Zero sales on a stockout day is missing data, not zero demand. Exclude those days. (Honey Shelf does this correction automatically; it's the single biggest reason its recommendations differ from a naive spreadsheet's.)
2. Ignoring supplier variability
"Lead time: 15 days" is a quote, not a fact. If your last six POs arrived in 14, 16, 21, 15, 19, and 17 days, planning on 15 means you stock out every time the supplier lands in the top half of their own distribution. Use the average as your lead time and push the variability into safety stock — and track promised-vs-actual on every PO so the numbers stay honest. A supplier scorecard with rolling on-time-delivery rates tells you when a partner's 15 days quietly became 20.
3. One reorder point for the whole product
A kurta that sells 4.2/day overall might sell 1.8/day in Medium, 1.1 in Large, 0.8 in Small, and 0.5 in XL. A single product-level ROP will stock out Medium weeks before the "product" looks low, while XL gathers dust. Calculate per variant. Yes, that's 4× the spreadsheet rows — which is usually the moment brands stop maintaining the spreadsheet and automate instead.
A reorder point tells you when — not how much
The ROP fires. What quantity do you order? A simple, robust answer is a target cover period: order enough to cover your lead time plus your review cycle — 45-60 days of velocity is a common range for made-to-order apparel. Then sanity-check it against three constraints: your manufacturer's minimum order quantity pushes the number up, your cash position pulls it down, and shelf life caps it hard for anything perishable. Honey Shelf pre-fills draft production orders with a recommended quantity computed from current velocity, so the when and the how-much arrive in the same click.
Manual vs automated reorder points
| Spreadsheet ROP | Automated ROP (Honey Shelf) | |
|---|---|---|
| Velocity source | Manually exported, quickly stale | Pulled from Shopify daily |
| Stockout-day correction | Almost never done | Automatic on every SKU |
| Recalculation | Monthly, if someone remembers | Daily, per variant |
| Lead-time input | Supplier's quoted number | Observed lead times per supplier |
| When ROP is crossed | Someone has to notice | Days-remaining alert + draft production order |
The formula is identical in both columns. The difference is whether it runs on fresh data, and whether crossing the line triggers an action or waits to be noticed. If stockouts keep happening despite "having reorder points," the process is the problem — our guide on avoiding stockouts covers the other eight strategies that surround this one.
Start with your top five SKUs
Don't boil the ocean. Pull 90 days of sales for your five best sellers, strike out the stockout days, compute velocity, look up your last few actual lead times, and set five reorder points this week. That alone typically prevents the stockouts that cost the most. Then extend to variants, then to the long tail — or let software do the extending for you.