Reorder Points for a Small Store: When to Order Without Overstocking

Every store owner knows both expensive mistakes. Order too late and the shelf goes empty — lost sales, disappointed regulars, rush orders at bad prices. Order too early or too much and cash sits in the back room — money that could have paid a bill, plus the risk that slow movers go stale before they sell.

A reorder point is the simple rule that keeps you between those two mistakes: a number per SKU that says “when the stock falls to this level, place the next order.” It is not software, and it is not a forecast of the future. It is arithmetic on three things you can observe: how fast the item sells, how long the supplier takes, and how much buffer you want against surprises.

The policy this guide uses (read this first)

This guide teaches a continuous-review policy: you watch each SKU’s inventory position on an ongoing basis, and when the position reaches the reorder point, you order a fixed quantity Q for that SKU. This is the standard “reorder point system,” also called a fixed-order-quantity system (see Sources).

That is different from periodic review: checking stock only at fixed intervals (every Monday, say) and ordering up to a target level. Periodic review is a legitimate policy, but it is not this guide — nothing here tells you to review on a schedule or to order up to a target. The two policies are not mixed in this article.

What a reorder point is — and what it isn’t

A reorder point answers when to order. It does not answer how much to order — under this guide’s policy, “how much” is a fixed quantity Q you choose per SKU (see the fixed-quantity section below). The reorder point also does not predict demand spikes, fix an unreliable supplier, or correct wrong inventory numbers. It assumes two things are already true: your on-hand quantities are trustworthy (if they aren’t, investigate the mismatch first), and you know roughly how long your supplier takes (if you don’t, start documenting your deliveries).

When demand and lead time are both constant, the reorder point is simply demand during lead time (NC State SCRC). With variability — and real stores have it — you add safety stock:

Reorder point = (average daily demand × lead time in days) + safety stock

One more definition, and it matters: you compare the reorder point against your inventory position, not just what’s on the shelf. Operations texts define it as:

Inventory position = on-hand (OH) + scheduled receipts (SR) − backorders (BO)

where scheduled receipts are orders placed but not yet received — also called open orders — and backorders are customer orders that cannot be filled when demanded but filled later (Krajewski et al., 2016, via Aslan 2023). Never reorder on on-hand alone — you will double-order while purchase orders are still in transit.

Three precision notes on this formula, because they prevent real errors:

  • On-hand means sellable stock. Damaged, expired, or already-sold units sitting in the back do not count. If your count mixes sellable and unsellable, your position is overstated.
  • On-order means confirmed open purchase orders. A verbal “yeah, I’ll send it” from a rep is not a scheduled receipt until the order is placed and confirmed.
  • Don’t deduct backorders twice. If your system already shows an “available” quantity with customer backorders netted out, use that number as your position — subtracting backorders again double-counts the deduction. Deduct only when your on-hand figure has not already been reduced.

And one limitation, stated plainly: inventory position does not reveal stockout risk before a late order arrives. If an open order is running two weeks late, your position still counts those units as coming — the math looks fine while the shelf is heading for empty. Position tells you when to order; it does not tell you an order is late. Watch confirmed arrival dates separately.

The three inputs, and where each comes from

1. Average daily demand. Pull units sold per SKU from your POS or ledger over a recent representative period — 90 days is a practical window — and divide. If you sold 540 units in 90 days, average daily demand is 6. Two cautions:

  • Units sold can understate true demand. Every unit you would have sold during a stockout never appears in the sales record. If the item stocked out during your window, your average is biased low — note the stockout days and adjust upward, or pick a cleaner window.
  • Document the period and any adjustment. Write down which dates you used and why, and what you adjusted. “90 days, Mar–May, excluded the one bulk sale on Apr 12” is a record; “about 6 a day” is a guess. Future reviews need the record.

2. Lead time in days. How long from placing the order to the goods being on your shelf and sellable — not just the supplier’s quoted transit time. Include your own processing: a supplier that ships in 7 days but whose deliveries you check in over 2 more days has a 9-day lead time for your purposes. Use your documented delivery history, not the brochure promise; suppliers quote averages and deliver realities.

Day convention (pick one, never mix): use calendar days for both demand and lead time throughout. If your store closes certain days and you prefer business days, use business days for both — average daily demand and lead time in the same units. Mixing calendar-day demand with business-day lead time silently breaks the multiplication. (ASCM: it is critical that the same time units are used for all variables.)

3. Safety stock. Extra units held against the two things you can’t schedule: demand running hotter than average, and deliveries arriving later than usual. NC State’s Supply Chain Resource Cooperative frames it exactly this way: safety stock is “additional inventory beyond amount needed to meet ‘average’ demand during lead time,” protecting “against uncertainties in demand or lead time” — and it balances the cost of stocking out against the cost of holding extra inventory. Two practical rules of thumb — labeled as what they are, simplifications, not statistical standards:

  • Days-of-cover: safety stock = average daily demand × N days, where N reflects how much a stockout hurts. A 3-day cover on a 6-unit/day item is 18 units.
  • Max-minus-average: safety stock = (maximum daily demand − average daily demand) × lead time. If the item averages 6/day, peaks at 10/day, and lead time is 9 days: (10 − 6) × 9 = 36 units. This sizes the buffer to the spikes you’ve actually seen.

Limits of max-minus-average, stated explicitly: it is valid only when you treat lead time as fixed — it says nothing about late deliveries. It is a rough rule that assumes the maximum daily demand persists through the entire lead time, which is a pessimistic assumption, not a measured one. It does not cover lead-time variability, and it achieves no specific service level — unlike statistical methods, where a Z-factor maps to a chosen cycle service level (NC State: Z ≈ 1.29 → 90%, 1.65 → 95%, 2.33 → 99%). If you need a defined service level, this rule of thumb is not the tool.

Pick one method per SKU and write down which — and record the real constraints alongside it: cash (how much capital the buffer ties up), expiry (perishables and dated goods cap any buffer), and space (shelf and backroom capacity are hard limits). A buffer that doesn’t fit in the store or dies on the shelf is not protection. ASCM’s practice guidance agrees with differentiating by SKU: set protection independently for product groups based on strategic importance, profit margin, and dollar volume — more safety stock where the money and the stockout cost are, less where they aren’t.

Worked example (hypothetical, arithmetic verified)

Everything in this section is hypothetical — made-up numbers for an imaginary store, shown so you can check the arithmetic. No field results are claimed.

Take SKU C-118, all-purpose cleaner 1L. Calendar days throughout:

  • Units sold, last 90 days: 540 (no stockout days in the window — stated as an assumption of the example) → average daily demand = 540 ÷ 90 = 6 units/day
  • Lead time (order to shelf-ready, from delivery records): 9 days
  • Lead-time demand = 6 × 9 = 54 units
  • Safety stock (max-minus-average): max daily demand 10, average 6 → (10 − 6) × 9 = 36 units
  • Reorder point = 54 + 36 = 90 units

Now check inventory position on a Tuesday morning: on-hand 55 sellable + on-order 30 confirmed − committed 0 = 85. Since 85 ≤ 90, place the order today. (Whether the open 30 covers the gap until the new order arrives depends on its confirmed arrival date — check it; the formula assumes it arrives on time.)

A second, quicker check — SKU T-204, trash bags: average daily demand 2, lead time 14 days → lead-time demand = 2 × 14 = 28; safety stock (5 days of cover) = 2 × 5 = 10; reorder point = 38. Position 41 → wait; position 37 → order.

How much to order: a fixed quantity Q (hypothetical illustration)

Under this guide’s policy, the reorder point says when; Q — a fixed quantity per SKU — says how much, and the same Q is ordered every time the trigger fires (IGI/Krajewski: “when the inventory position reaches R, a fixed quantity Q of the SKU is ordered”). Choosing Q is a management decision, not a formula output of this guide. A practical way to choose it: pick days of supply you can afford and fit.

Continuing the hypothetical: C-118 sells 6/day. You choose 180 units as the fixed order quantity — a hypothetical Q chosen to cover about 30 days of demand (6 × 30 = 180), rounded to the supplier’s case pack: cases of 12 → 180 ÷ 12 = 15 cases. To be explicit about what this is and isn’t: it is a fixed Q you selected; it is not a computed order-up-to quantity, not derived from a target stock level, and not a guarantee that you will order monthly — the actual time between orders varies with demand, because orders fire when position hits the reorder point, not on a calendar.

Two constraints that can override your chosen Q:

  • Supplier minimum (MOQ). If your chosen Q falls below the supplier’s minimum order quantity, that is a decision point, not an automatic order: raise Q, drop the SKU, find another supplier, or combine the order with other SKUs — decide deliberately.
  • Budget, expiry, space. The same three constraints from safety stock apply to Q: don’t tie up cash you need, don’t order perishables past their sellable life, don’t order what won’t fit.

(Formal order-quantity optimization — EOQ and its variants — exists, but it needs ordering-cost and holding-cost inputs most small stores haven’t measured; the days-of-supply starting point above is the honest beginning, not a substitute.)

Five mistakes that break reorder points

  1. Reordering on on-hand alone. The classic double-order: 55 sellable on the shelf looks low, but 30 confirmed are in transit. Always use inventory position.
  2. Using the supplier’s quoted lead time. Quotes are averages on good weeks. Use your documented history — and if the supplier is erratic, say so and carry more safety stock rather than pretending the average is reliable. (The two factors that spoil the textbook reorder point are the fluctuation of demand and of lead time — Aslan 2023.)
  3. Setting it once and forgetting it. Demand drifts, suppliers change, seasons turn. Review reorder points quarterly — or whenever a stockout or overstock surprises you, which is the system telling you an input moved.
  4. Building on bad numbers. A reorder point computed from a wrong on-hand quantity confidently orders the wrong amount. If counts and system disagree, investigate before you trust the numbers — and keep the counts honest with a regular cycle-count routine.
  5. One rule for every SKU. Differentiate deliberately: more protection where strategic importance, margin, and dollar volume are high and stockouts are costly; thinner buffers elsewhere — and always within the hard limits of cash, expiry, and space. The ABC thinking from your counting routine applies here too: spend the precision where the money is.

The one-page worksheet

Print one per review cycle and work SKU by SKU: SKU/item, average daily demand, lead time (days), lead-time demand, safety stock, safety-stock method used (per SKU), reorder point, on-hand (sellable), on-order (confirmed), committed, inventory position, order now? (Y/N), fixed order quantity Q. The formulas and the corrected definitions are printed on the sheet. Download the reorder-point worksheet (HTML, prints to 1 page on US Letter).

FAQ

Do I need inventory software for this?

No. The math fits on the worksheet: three inputs per SKU, one multiplication and one addition. Software earns its place when you have hundreds of SKUs and the arithmetic itself becomes the bottleneck — the same complexity caution applies here as with counting: buy software when the paper proves too slow, not before.

What if demand is seasonal?

Then a single annual average lies. Split the year: compute separate reorder points for the season you’re entering, using last year’s same-season sales as the demand input — and write the season and the dates used on the worksheet so next year’s review starts from the right numbers. If you don’t have last year’s data, say so, use the recent average, and carry extra safety stock until you do. (ASCM: if seasonality is treated as normal variability, you can end up with very high safety stock yet still run short in the peak season.)

My supplier’s lead time is all over the place. What then?

Two honest options: carry safety stock sized to the worst typical case (not the average), or find a more reliable supplier for the SKUs where stockouts hurt most. What doesn’t work is using the quoted average and hoping — hope is not a buffer. Note the boundary: max-minus-average covers demand spikes under a fixed lead time; it does not cover lead-time variability.

Sources (all read directly, 2026-10-11)

  1. NC State University, Supply Chain Resource Cooperative — “SAFETY STOCK ANALYSIS: Inventory Management Models: A Tutorial.” https://scm.ncsu.edu/scm-articles/article/safety-stock-analysis — Used for: reorder point = demand during lead time under constant conditions; safety stock as inventory beyond average lead-time demand protecting against demand/lead-time uncertainty; balancing stockout vs. holding cost; Z ≈ 1.29 → 90%, 1.65 → 95%, 2.33 → 99% cycle service level.
  2. ASCM Insights — Peter L. King & Courtney Bigler, “Calculate Inventory with Precision — Even Amid Variability.” https://www.ascm.org/ascm-insights/safety-stock-a-contingency-plan-to-keep-supply-chains-flying-high/ — Used for: safety stock protects against forecast errors and demand fluctuations; correct level depends on supply and demand variability; same time units required for all variables; differentiate protection by product group (strategic importance, margin, dollar volume); handle seasonality via cycle stock, not inflated safety stock.
  3. IGI Global — Emre Aslan (2023), “Determination of Inventory System Service Level by Simulation…” (citing Krajewski et al., 2016). https://www.igi-global.com/chapter/determination-of-inventory-system-service-level-by-simulation-in-cases-of-constant-and-variable-demand-and-lead-time/326345 — Used for: continuous review = fixed order quantity = reorder-point system; when inventory position reaches R, a fixed quantity Q is ordered; IP = OH + SR − BO (SR = orders placed but not received; BO = customer orders unfilled now, filled later); demand and lead-time fluctuation spoil the textbook case.