Reorder point and safety stock calculator (with EOQ)
Reorder point = average usage during the supplier lead time + safety stock. Safety stock = z × σ of usage per period × √(lead time in periods) when the lead time is fixed, and z × √(L·σd² + d̄²·σL²) when it varies; z is 1.6449 for 95% and 2.3263 for 99% of order cycles without a stock-out. Example (Chopra, Reinhardt & Dada, 2004): 2,500 units a week with σ 500, a 2-week lead time and 95% give safety stock 1,164 and a reorder point of 6,164. Order quantity: EOQ = √(2DS/H). In Logistivo, the reorder point goes into a product’s Min. Stock field and the order quantity into Reorder Qty; a product is flagged Low once its stock on hand across all warehouses falls below Min. Stock.
Source: APICS (now ASCM) — Peter L. King · S. Chopra, G. Reinhardt, M. Dada — Kellogg School of Management · MIT OpenCourseWare (CC BY-NC-SA) · ACCA · NIST/SEMATECH e-Handbook of Statistical Methods · M. J. Wichura — StatLib, Carnegie Mellon University · F. W. Harris — INFORMS · E. A. Silver, D. F. Pyke, D. J. Thomas — CRC Press · Microsoft Support · legislation.gov.uk · Financial Reporting Council · IFRS Foundation.
Data last updated:
.
The cycle, drawn at average usage
The cycle, drawn at average usage — Fill in the sentence — get when to reorder, how much buffer to hold and how much to order.
The calculation runs in your browser; nothing you type is sent or stored. It assumes normally distributed demand during the lead time, continuous review unless you enter a review period, and demand independent of the lead time unless you choose “move together”.
We use [Average usage per period] per [Data period], give or take [Usage variability (standard deviation per period)]. The supplier delivers in [Supplier lead time, days] days, give or take [Lead-time variability, days (standard deviation)]. We want [Cycle service level, %]% of order cycles to end without a stock-out.
Order quantity and fine-tuning — We order [Order quantity (optional)] at a time — or work it out with EOQ from [Annual demand, units (optional, for EOQ)] a year, [Cost per order (optional, for EOQ)] per order and [Holding cost per unit per year (optional, for EOQ)] to hold one unit for a year. We check stock every [Review period, days (optional)] days (leave it empty if you check continuously); demand and lead time [Demand and lead time]; we count in [Units].
Average usage per period — Mean demand in the period you choose: day, week or month. Usage and its variability must describe the same period.
Data period — The lead time is converted into this period: σ over the lead time = σ per period × √(lead time ÷ period) (APICS, 2011).
Usage variability (standard deviation per period) — Leave 0 if usage is steady, or paste your history to compute it. Sample standard deviation with n − 1 (NIST; Excel STDEV.S).
Supplier lead time, days — From placing the order to the goods being usable in stock. Counted in the same kind of day as the period: calendar days with a 7-day week or a 30-day month, working days with a 5-day week.
Lead-time variability, days (standard deviation) — 0 if deliveries always take the same time.
Cycle service level, % — Share of order cycles that end without a stock-out; 50 to 99.9%. z = Φ⁻¹(CSL), computed with Wichura’s AS241 algorithm (1988).
Order quantity (optional) — Needed for the chart and the fill rate. Left empty, EOQ is used when D, S and H are filled in.
Annual demand, units (optional, for EOQ) — Never prefilled: EOQ is only computed from your own figures.
Review period, days (optional) — If stock is checked only every R days, the calculator uses L + R in both terms.
Demand and lead time — Independent: the variances are added under one square root. Moving together: the two safety stocks are added (APICS, 2011).
Units — Whole units are rounded up; divisible units keep three decimals, as Logistivo’s quantity fields do.
Usage history (optional) — Paste 8 to 30 daily or weekly figures separated by spaces, line breaks or semicolons, then use them. The calculator fills in the average and the sample standard deviation (n − 1) and shows the coefficient of variation.
Data period — day · week (7 days) · working week (5 days) · month (30 days)
Demand and lead time — move together · vary independently
Units — whole units · kg, litres or other divisible units
Reorder at — Enter this in Min. Stock in Logistivo
Safety stock — The part of the reorder point that absorbs variability
Order quantity — As typed — enter it in Reorder Qty · From EOQ — enter it in Reorder Qty · Add a quantity, or D, S and H below
Load a worked example from the sources
Chopra et al., 2004 — 2,500 a week, σ 500, 2-week lead time, orders of 10,000
MIT OCW, lecture 5 — 44.6 a week, σ 32.1, 2 weeks, 97% → about 175
APICS 2011, film rolls — 50 rolls a week, σ 10, 8 days ± 0.5, 95% → 19 rolls
ACCA EOQ example — 60,000 a year, $27 per order, $1 to hold → 1,800
MIT: steady item A — 8 weekly figures, 1-week lead time, 97% → s = 65 — 33 45 37 38 55 30 18 58
Drawn at average usage from the moment an order arrives. Real usage wobbles around the line; the safety-stock band is what absorbs it. The dashed grey line is the inventory position — stock on hand plus orders placed but not yet delivered — and it is this line, not stock on hand, that meets the reorder point.
Enter an order quantity, or annual demand, cost per order and holding cost for EOQ, to draw the cycle. The calculator never assumes one.
At this order quantity more than 300 orders would be in transit at once, so the chart is not drawn. Check the quantity and the period.
Safety stock · Reorder point · Lead time: order placed → goods in · Stock review · Stock on hand · Inventory position (on hand + on order)
How your numbers were built — Each rung is one term of the formula with your figures substituted. Change the sentence above and every rung recalculates.
L ÷ period — The lead time — plus the review period, if any — is expressed in the period of your usage data before anything is multiplied. — APICS (2011): σ is taken over periods equal to the total lead time, including any review period.
d̄ × L — Average usage during the lead time: what an ordinary cycle consumes while the order is on its way. — MIT OCW ESD.273J lecture 5 (LT × AVG); ACCA: re-order level = demand in lead time.
z = Φ⁻¹(CSL) — The service factor: how many standard deviations of lead-time demand the buffer covers. Computed exactly, not looked up. — Wichura, AS241 (1988); agrees with NIST’s table to every printed decimal.
σ over the lead time — Fixed lead time: σd × √L. Variable and independent: √(L × σd² + d̄² × σL²). Moving together: σd × √L + d̄ × σL. — Chopra, Reinhardt & Dada (2004), citing Silver & Peterson (1985); APICS (2011).
lead-time share — How much of the uncertainty comes from the supplier’s timing rather than from usage: d̄² × σL² ÷ (L × σd² + d̄² × σL²). It shows which variability to work on first. — APICS film-roll example: demand variability’s effect was almost 10 times lead time’s (share about 10%).
SS = z × σ — Safety stock, rounded up to a whole unit for counted items. — APICS rounds 17.6 to 18 rolls and 18.5 to 19.
ROP = d̄ × L + SS — The reorder point: when the inventory position falls to this level, place the order. — MIT OCW lecture 5; Chopra et al. (2004): ROP = M + zS.
EOQ = √(2DS ÷ H) — Economic order quantity from annual demand D, cost per order S and holding cost per unit per year H. — Harris (1913); ACCA worked example √(2 × 27 × 60,000 ÷ 1) = 1,800.
β = 1 − σ × G(z) ÷ Q — Fill rate: the share of demand served straight from stock; G(z) is the standard normal loss function. It needs an order quantity; when the expected shortage per cycle exceeds the order quantity the approximation no longer holds and no rate is shown. — Reproduces Chopra et al. (2004), Table 2: safety stock 0 → cycle service level 50%, fill rate 97.18%.
Checks the calculator runs
Average usage must be above zero.
The usage standard deviation cannot be negative.
The lead time must be above zero days.
Lead-time variability cannot be negative.
A 100% cycle service level is statistically unattainable (APICS, 2011): the normal curve never reaches zero. The calculator accepts up to 99.9%.
Below 50% the z-value turns negative and the formula returns negative safety stock (Chopra et al., 2004, Figure 1). The calculator accepts 50 to 99.9%.
Enter a cycle service level between 50% and 99.9%.
With no variability in usage or lead time, safety stock is zero and the reorder point equals lead-time demand.
The coefficient of variation is above 0.40 — outside the range of the empirical study cited by Chopra et al. (Tyworth & O’Neill, 1997: coefficients of variation below 40%). For lumpy items, such as MIT’s item B with CV 1.21, treat the result as a starting point.
Lead-time usage exceeds one order quantity, so more than one order is in transit when stock reaches the reorder point. Compare the reorder point with the inventory position (on hand + on order), not with stock on hand (MIT OCW lecture 5).
With a review period the stock has to last until the next review plus the lead time, so the calculator uses L + R in both terms and the reorder point applies at each review.
Moving together: the two safety stocks are added instead of combined under one square root, which gives more stock (APICS, 2011).
Most of the uncertainty comes from the supplier’s lead time. Chopra et al. (2004) find that at 95% reducing lead-time variability lowers safety stock more than shortening the lead time, but at the 50–70% cycle service levels they say most firms actually run, shortening the lead time wins.
Formulas and rules the calculator applies
Reorder point — ROP = d̄ × L + SS — MIT OCW ESD.273J lecture 5; Chopra et al. (2004)
Safety stock, fixed lead time — SS = z × σd × √L — APICS (2011)
Safety stock, variable lead time (independent) — SS = z × √(L × σd² + d̄² × σL²) — Chopra et al. (2004), citing Silver & Peterson (1985); APICS (2011)
Safety stock, variability moving together — SS = z × σd × √L + z × σL × d̄ — APICS (2011)
Different time units — σ over the lead time = σ per period × √(lead time ÷ period); weekly data and a 3-week lead time → σ per week × √3 — APICS (2011)
Periodic review — replace L with L + R — APICS (2011): “total lead time, including any review period”
Fill rate — β = 1 − σ × G(z) ÷ Q, with G(z) = φ(z) − z × (1 − Φ(z)) — Chopra et al. (2004), Table 2 reproduced
Economic order quantity — EOQ = √(2 × D × S ÷ H) — Harris (1913); ACCA
Bounds — cycle service level 50 to 99.9%; 100% is unattainable; below 50% safety stock turns negative — APICS (2011); Chopra et al. (2004)
Rounding — whole units: safety stock and reorder point are each rounded up from their exact values; divisible units: three decimals — APICS (2011) rounds 17.6 → 18 and 18.5 → 19
Lumpy-demand warning — coefficient of variation σd ÷ d̄ above 0.40 — Tyworth & O’Neill (1997), as cited by Chopra et al. (2004)
Standard deviation and z in Excel — sample standard deviation with n − 1: STDEV.S; z: =NORM.S.INV(0.95) returns 1.6449 — NIST 1.3.5.6; Microsoft Support
Assumptions
Demand during the lead time is normally distributed.
Demand is independent of the lead time unless you choose “move together”.
Stock is reviewed continuously unless you enter a review period.
The fill rate assumes unmet demand is backordered, not lost.
EOQ assumes a constant purchase price, constant demand and lead time, a holding cost that depends on average inventory and an order cost that does not depend on the quantity; quantity discounts need a separate total-cost comparison (ACCA).
The chart shows average usage: it is a picture of the rule, not a forecast.
A planning aid built on published formulas. Check the result against your own stock history before you change a setting that drives purchasing.
How do I calculate a reorder point?
How do I calculate a reorder point? — Multiply average usage by the supplier lead time, in the same time unit, and add safety stock: ROP = d̄ × L + SS. Order when the inventory position — stock on hand plus anything already ordered — falls to that level.
d̄ × L — Lead-time demand: what an average cycle uses while the order travels. — Worked example: MIT’s distributor sells 44.6 units a week with a 2-week lead time: 44.6 × 2 = 89.2. — MIT OCW ESD.273J, lecture 5 (D. Simchi-Levi)
+ SS — Safety stock covers usage above the average during that wait. — Worked example: At 97% (z = 1.88) with σ 32.1 a week: 1.88 × 32.1 × √2 = 85.3, so the reorder point is about 175 — “about 3.9 weeks of supply at warehouse and in the pipeline”. — MIT OCW ESD.273J, lecture 5
inventory position — Compare the reorder point with stock on hand plus stock on order, not with stock on hand alone; otherwise the next check orders a second time for the same shortfall. — Worked example: Chopra’s product (14-day lead time, 28-day cycle) has one order in transit at a time; APICS’s film rolls (8-day lead time, weekly batches) have two. — MIT OCW lecture 5: inventory position is “the actual inventory plus items already ordered, but not yet delivered”.
UK wording — UK management-accounting texts call the reorder point the re-order level (ROL) and safety stock buffer inventory. — Worked example: ACCA: with certain demand and lead time, “re-order level (ROL) = demand in lead-time”; buffer inventory “is simply the amount by which ROL exceeds average demand in lead-time”. — ACCA Student Accountant, “Inventory control”
What is the safety stock formula, and when do I need the lead-time-variability version?
What is the safety stock formula, and when do I need the lead-time-variability version? — When the supplier’s lead time is reliable, SS = z × σd × √L. As soon as deliveries vary, add the lead-time term: SS = z × √(L × σd² + d̄² × σL²). If late deliveries and demand spikes tend to happen together, add the two safety stocks instead.
z × σd × √L — Fixed lead time: only usage varies. — Worked example: 2,500 a week, σ 500, 2 weeks, 95%: 1.6449 × 500 × √2 = 1,163.09 → 1,164. — APICS (2011); parameters of Chopra et al. (2004), Table 2
z × √(L·σd² + d̄²·σL²) — Variable lead time, independent of demand: the two variances add, then take the square root. — Worked example: APICS film rolls: 50 a week, σ 10, 8-day cycle, lead-time σ half a day (0.07 weeks): 1.65 × √(8/7 × 10² + (0.07 × 50)²) = 19 rolls, against 18 with a fixed lead time. — APICS (2011); Chopra et al. (2004), citing Silver & Peterson (1985)
z·σd·√L + z·σL·d̄ — When demand and lead-time variability are not independent — both driven by the same cause — the square-root form understates the risk; add the two safety stocks. — Worked example: The same film rolls, moving together: 1.6449 × 10 × √(8/7) + 1.6449 × (0.5 ÷ 7) × 50 = 17.58 + 5.87 = 23.46 → 24 rolls. — APICS (2011)
which term dominates? — Compare d̄² × σL² with L × σd²: the larger one is where improvement pays. — Worked example: In the film-roll case demand variability’s effect is “almost 10 times that of lead time variability”, a lead-time share of 10%. The ladder above shows your share. — APICS (2011)
Which service level should I use: 95% or 99%?
Which service level should I use: 95% or 99%? — Most businesses set cycle service levels between 90% and 98%; 100% is statistically unattainable. Moving from 95% to 99% raises z from 1.6449 to 2.3263 — 41% more safety stock for the same variability — so set it per product group rather than one figure for the whole catalogue.
90–98% — Typical goals fall in this range. — Worked example: APICS: “Typical goals fall between 90 and 98 percent”. — APICS (2011)
per product group — Set z per product group — strategic importance, profit margin or sales value — so the items that matter most carry more buffer. — Worked example: 97.5% (z 1.9600) against 90% (z 1.2816): at the same σ the 90% group needs 35% less safety stock. — APICS (2011)
95% → 99% — The relationship is not linear: each extra point of service costs more stock than the one before. — Worked example: 2.3263 ÷ 1.6449 = 1.414, i.e. 41% more safety stock; with Chopra’s data 1,164 → 1,645 units. — Computed with AS241
100% — No finite buffer guarantees every cycle. — Worked example: The calculator accepts 50 to 99.9% (z up to 3.0902). — APICS (2011)
Service level z-score table (with the loss function G(z))
Service level z-score table (with the loss function G(z)) — z = Φ⁻¹(CSL), computed with Wichura’s AS241 algorithm; G(z) = φ(z) − z × (1 − Φ(z)) is used for the fill rate. z = 1 corresponds to 84.13%.
50 — z: 0.0000 | G(z): 0.39894
75 — z: 0.6745 | G(z): 0.14915
80 — z: 0.8416 | G(z): 0.11164
85 — z: 1.0364 | G(z): 0.07769
90 — z: 1.2816 | G(z): 0.04734
92 — z: 1.4051 | G(z): 0.03626
95 — z: 1.6449 | G(z): 0.02089
96 — z: 1.7507 | G(z): 0.01615
97 — z: 1.8808 | G(z): 0.01162
97.5 — z: 1.9600 | G(z): 0.00945
98 — z: 2.0537 | G(z): 0.00734
99 — z: 2.3263 | G(z): 0.00339
99.5 — z: 2.5758 | G(z): 0.00158
99.8 — z: 2.8782 | G(z): 0.00058
99.9 — z: 3.0902 | G(z): 0.00028
NIST’s standard normal table gives 1.282, 1.645, 1.960, 2.326, 2.576 and 3.090 for 90, 95, 97.5, 99, 99.5 and 99.9% — the same values rounded to three decimals.
In Excel: =NORM.S.INV(0.95) returns 1.6449. Microsoft’s own example: =NORM.S.INV(0.908789) returns 1.3333347.
The row that matches the service level in your sentence is highlighted.
Is a 95% service level the same as a 95% fill rate?
Is a 95% service level the same as a 95% fill rate? — No. Cycle service level counts order cycles that end without a stock-out; fill rate counts the share of demand served from stock. With zero safety stock a product can still fill 97% of demand while stocking out in half of its cycles.
cycle service level — The probability that a replenishment cycle has no stock-out — the percentage z is built on. — Worked example: Zero safety stock → z = 0 → 50% of cycles without a stock-out. — APICS (2011); Chopra et al. (2004)
fill rate β — β = 1 − σ × G(z) ÷ Q: the expected shortage per cycle divided by the order quantity. — Worked example: 2,500 a week (σ 500), 2 weeks, orders of 10,000: safety stock 0 → CSL 50.0%, fill rate 97.18%; 200 → 61.1%, 98.07%; 400 → 71.4%, 98.74%. — Chopra et al. (2004), Table 2 — the calculator reproduces all three rows
97–99% fill rate — Chopra et al. argue that most firms aim for fill rates of 97–99%, which on data like theirs means cycle service levels of only 50–70%. — Worked example: At those levels the normal approximation is most distorted, which is why they advise shortening the lead time rather than chasing lead-time variability. — Chopra et al. (2004)
which one to report — Fill rate is often the better performance measure, because cycle service level shows how often you stock out but not how much. — Worked example: With stable demand and lead time the fill rate runs above the cycle service level; with volatile ones it can fall below it. — APICS (2011), Figure 3
My usage data are weekly but the lead time is in days: what changes?
My usage data are weekly but the lead time is in days: what changes? — Convert before you multiply. The standard deviation grows with the square root of time, so σ over the lead time = σ per period × √(lead time ÷ period). Mixing a weekly σ with a lead time in days skews safety stock by that factor.
√(lead time ÷ period) — The lead time here is the total lead time including any review period; the period is the one your data describe. — Worked example: Weekly σ and a 3-week lead time → σ per week × √3. Film rolls: 8-day cycle and weekly data → × √(8/7) = × 1.069. — APICS (2011)
same kind of day — Count the lead time in the same kind of day as the data period. — Worked example: Calendar days go with a 7-day week or a 30-day month, working days with a 5-day week; the calculator’s period menu states which one it uses.
review period — If you only check stock every R days, the stock has to last R + L. — Worked example: Weekly checks and a 14-day lead time: protect 21 days, not 14. — APICS (2011)
sample σ — Estimate σ from history with the sample standard deviation, which divides by n − 1. — Worked example: Excel STDEV.S, not STDEV.P. MIT’s item A over 8 weeks: mean 39.25, s 13.18, coefficient of variation 0.34. — NIST 1.3.5.6; Microsoft Support; MIT OCW lecture 5
How is EOQ calculated, and what does it assume?
How is EOQ calculated, and what does it assume? — EOQ = √(2 × D × S ÷ H), with D the annual demand, S the cost of placing one order and H the cost of holding one unit for a year. ACCA’s example — 60,000 units a year, $27 per order, $1 to hold — gives 1,800 units.
√(2DS ÷ H) — The quantity at which total ordering plus holding cost is lowest. — Worked example: √(2 × 27 × 60,000 ÷ 1) = 1,800. — ACCA; Harris (1913)
assumptions — A constant purchase price, constant demand and lead time, a holding cost that depends on average inventory and an order cost independent of the order quantity. — Worked example: A supplier’s price break breaks the first assumption: compare total annual cost at the EOQ and at the discount quantity instead. — ACCA
how much vs when — EOQ sets how much to order; safety stock and the reorder point set when. They are separate settings. — Worked example: Logistivo keeps them apart too: Reorder Qty for the quantity, Min. Stock for the trigger. — Logistivo stock module
Harris, 1913 — The formula goes back to F. W. Harris, “How Many Parts to Make at Once”. — Worked example: Reprinted in Operations Research 38(6), 1990. — Harris (1913)
What should I enter as the minimum stock level in inventory software?
What should I enter as the minimum stock level in inventory software? — Enter the reorder point, not the safety stock. A minimum-stock alert exists to tell you when to order; set to the safety stock it fires only after the lead-time demand is gone and you are already using the buffer.
Min. Stock = ROP — The minimum-stock field is the trigger level. — Worked example: Chopra’s data at 95%: Min. Stock 6,164 — not 1,164. — Follows from ROP = d̄ × L + SS
Reorder Qty = Q — The reorder quantity holds how much to order: your EOQ, or the supplier’s pack or lot size. — Worked example: ACCA’s example: Reorder Qty 1,800. — Logistivo stock module
“Low” — In Logistivo a product is flagged Low when its total stock on hand across all your warehouses is below Min. Stock; the Products and Stock Levels tabs can be filtered to low stock only. — Worked example: Min. Stock 6,164 with 6,100 on hand across two warehouses → Low; exactly 6,164 on hand → not Low, because the rule is “below”. — Logistivo stock module
open orders — The badge compares stock on hand, so after you place an order it stays on until the goods are booked in. — Worked example: Keep open orders in view before re-ordering: the reorder point is meant for the inventory position, on hand plus on order. — MIT OCW lecture 5
Warehouse management in Logistivo — /en/warehouse-management
When does the normal safety stock formula mislead?
When does the normal safety stock formula mislead? — When usage is lumpy or the lead time is skewed. The formulas assume normally distributed demand during the lead time; Chopra et al. show that with skewed lead times the error can run in either direction.
CV above 0.40 — The empirical support for the normal approximation comes from fast movers with coefficients of variation below 40%. — Worked example: MIT’s item B over 8 weeks — 0 2 3 0 0 1 3 0 — has mean 1.125, s 1.36, CV 1.21. The calculator flags any CV above 0.40. — Tyworth & O’Neill (1997), as cited by Chopra et al. (2004); MIT OCW lecture 5
skewed lead time — With gamma-distributed lead times at 95%, the normal approximation understated safety stock. — Worked example: Lead time mean 10, σ 5: normal approximation 182 units, exact value 218. — Chopra et al. (2004), Table 1
service level 50–70% — At lower service levels the approximation points to the wrong lever. — Worked example: At 60%, cutting lead-time σ from 5 to 4 raised the exact safety stock from 20 to 22, while cutting the lead time from 10 to 8 lowered it to 15. — Chopra et al. (2004), Table 1
check it in use — Treat the formula as a starting point for lumpy or intermittent items and check how often the buffer is actually touched. — Worked example: APICS: is the safety stock being consumed in about half of the cycles? If not, look for special causes before changing the setting. — APICS (2011)
What stock records must a UK company keep, and how is stock valued?
What stock records must a UK company keep, and how is stock valued? — A company that deals in goods must keep statements of stock at each financial year-end and the stocktaking sheets behind them, for 3 years (private company) or 6 years (public company). Under FRS 102 stock is measured at the lower of cost and estimated selling price less costs to complete and sell.
Companies Act 2006 s.386(4) — Accounting records must contain statements of stock held at the end of each financial year, all statements of stocktakings from which they were prepared and, except for ordinary retail trade, statements of all goods sold and purchased. — What it means for stock: Year-end count sheets are part of the statutory records, not working papers. — legislation.gov.uk
s.388(4) — Those records are kept for 3 years from when they are made in a private company and 6 years in a public company. — What it means for stock: Keep the stocktake sheets for as long as the stock statements they support. — legislation.gov.uk
FRS 102, para 13.4 — Inventories are measured at the lower of cost and estimated selling price less costs to complete and sell. — What it means for stock: Safety stock that has stopped moving is exactly where this test bites. — FRS 102, September 2024 edition (periods beginning on or after 1 January 2026)
paras 13.18–13.19 — Cost is assigned by FIFO or weighted average — LIFO is not permitted — and inventories are assessed for impairment at each reporting date, for example for damage, obsolescence or declining selling prices. — What it means for stock: A buffer sized for a product that is being phased out is an impairment candidate at the next year-end. — FRS 102
IAS 2 — Companies reporting under IFRS measure inventories at the lower of cost and net realisable value. — IFRS Foundation
Common safety stock and reorder point mistakes
Setting safety stock as a flat 10–20% of cycle stock — APICS notes such rules “generally result in poor performance”.
Mixing periods: a weekly σ with a lead time in days, without the √(lead time ÷ period) factor.
Reading cycle service level as fill rate: at zero safety stock the first is 50%, the second can be 97%.
Using the fixed-lead-time formula when supplier lead times vary.
Combining the terms under a square root when demand and lead time move together, or adding them when they are independent.
Comparing the reorder point with stock on hand instead of the inventory position, which leads to double orders.
Typing the safety stock instead of the reorder point into a minimum-stock field.
Forgetting the review period when stock is only checked weekly or monthly.
Assuming normally distributed demand for lumpy, intermittent items.
Targeting 100% (unattainable) or entering a service level below 50% (negative safety stock).
Using EOQ where its assumptions fail, for example with supplier price breaks.
Calculating the population standard deviation (STDEV.P) instead of the sample one (STDEV.S, n − 1).
Related tools
Related tools — Lead time is the input most people guess. These two pages help pin it down and keep the result where the stock is.
Transport management system ETA calculator — Works out a legal lorry arrival time under the drivers’ hours rules — a transit-time input for the lead time above.
Warehouse management in Logistivo — Products, stock levels across warehouses, Min. Stock and the low-stock filter.
OTIF calculator — On-time and in-full delivery rate from your own delivery log, with the rule for “on time” and the tolerance you choose.
Where these two numbers live in Logistivo — The calculator is free. In Logistivo’s stock module the reorder point and the order quantity are fields on every product.
Min. Stock on every product — Enter the reorder point on the product; the same field is in the web panel and in the iOS and Android apps.
Low badge and filter — A product whose total stock on hand across all warehouses is below Min. Stock is flagged Low, and the Products and Stock Levels tabs filter to low stock only.
Reorder Qty next to the trigger — Store the order quantity — EOQ, pack size or supplier minimum — beside the reorder point.
Set it by asking — The Logistivo assistant can create a product or update one with both fields when you ask it.
Keep the reorder point where the stock is
Keep the reorder point where the stock is — Put Min. Stock and Reorder Qty on every product and let the Low filter show what to order. Plans and prices are explained in the pricing guide.
Source
APICS (now ASCM) — Peter L. King — Crack the Code: Understanding safety stock and mastering its equations (APICS magazine, July/August 2011) — https://web.archive.org/web/20131101135411/http://media.apics.org:80/omnow/Crack%20the%20Code.pdf
S. Chopra, G. Reinhardt, M. Dada — Kellogg School of Management — The Effect of Lead Time Uncertainty on Safety Stocks — Decision Sciences 35(1):1–24, 2004 (doi:10.1111/j.1540-5414.2004.02332.x) — https://web.archive.org/web/20241118192152/https://www.kellogg.northwestern.edu/faculty/chopra/htm/research/Effect%20of%20Lead%20Time%20Uncertainty.pdf
MIT OpenCourseWare (CC BY-NC-SA) — ESD.273J Logistics and Supply Chain Management, Fall 2009 — Lecture 5, Inventory management (D. Simchi-Levi) — https://ocw.mit.edu/courses/esd-273j-logistics-and-supply-chain-management-fall-2009/resources/mitesd_273jf09_lec05/
NIST/SEMATECH e-Handbook of Statistical Methods — 1.3.6.7.1 Cumulative distribution function of the standard normal distribution — https://www.itl.nist.gov/div898/handbook/eda/section3/eda3671.htm
NIST/SEMATECH e-Handbook of Statistical Methods — 1.3.5.6 Measures of scale — sample standard deviation (N − 1) — https://www.itl.nist.gov/div898/handbook/eda/section3/eda356.htm
M. J. Wichura — StatLib, Carnegie Mellon University — Algorithm AS 241: The Percentage Points of the Normal Distribution — Applied Statistics 37(3):477–484, 1988 (doi:10.2307/2347330) — https://lib.stat.cmu.edu/apstat/241
F. W. Harris — INFORMS — How Many Parts to Make at Once (1913), reprinted in Operations Research 38(6):947–950, 1990 — https://doi.org/10.1287/opre.38.6.947
E. A. Silver, D. F. Pyke, D. J. Thomas — CRC Press — Inventory and Production Management in Supply Chains, 4th edition (general reference) — https://doi.org/10.1201/9781315374406
Microsoft Support — NORM.S.INV function — https://support.microsoft.com/en-gb/excel/functions/norm-s-inv-function
Microsoft Support — STDEV.S function — https://support.microsoft.com/en-gb/excel/functions/stdev-s-function
legislation.gov.uk — Companies Act 2006, section 386 — duty to keep accounting records — https://www.legislation.gov.uk/ukpga/2006/46/section/386
legislation.gov.uk — Companies Act 2006, section 388 — where and for how long records to be kept — https://www.legislation.gov.uk/ukpga/2006/46/section/388
Financial Reporting Council — FRS 102, September 2024 edition — Section 13 Inventories — https://www.frc.org.uk/documents/7668/FRS_102_September_2024_tmKYWO6.pdf
IFRS Foundation — IAS 2 Inventories — https://www.ifrs.org/issued-standards/list-of-standards/ias-2-inventories/
Frequently asked questions
How do I calculate a reorder point?
Average usage per period × lead time in the same periods + safety stock. MIT’s example: 44.6 a week × 2 weeks = 89.2, plus 1.88 × 32.1 × √2 = 85.3 safety stock, gives a reorder point of about 175.
What is the safety stock formula?
z × σd × √L when the lead time is fixed; z × √(L·σd² + d̄²·σL²) when it varies independently of demand; if demand and lead time move together, add the two terms: z·σd·√L + z·σL·d̄ (APICS, 2011).
Which service level should I pick?
Typical goals are 90–98%, and 100% is unattainable (APICS). Set it per product group by importance, margin or value. Going from 95% to 99% takes 41% more safety stock for the same variability.
Is a 95% service level the same as a 95% fill rate?
No. Cycle service level counts cycles without a stock-out; fill rate counts demand served from stock. With 2,500 a week (σ 500), a 2-week lead time and orders of 10,000, zero safety stock gives a 50% cycle service level but a 97.18% fill rate (Chopra et al., 2004).
My data are weekly and the lead time is in days. What do I do?
Multiply the weekly σ by √(lead time in days ÷ 7), and add any review period to the lead time first (APICS, 2011). With working-day data, divide by 5 instead of 7.
How is EOQ calculated?
EOQ = √(2 × D × S ÷ H). ACCA’s example: √(2 × 27 × 60,000 ÷ 1) = 1,800. It assumes a constant price and constant demand and lead time; price breaks need a total-cost comparison.
What goes into “minimum stock” in inventory software?
The reorder point, not the safety stock — otherwise the alert fires only when you are already using the buffer. In Logistivo, Min. Stock flags a product Low once its total stock on hand across all warehouses drops below it.
Should I cut the lead time or its variability?
It depends on the service level. Chopra et al. (2004) find that at 95% reducing lead-time variability lowers safety stock more; at the 50–70% cycle service levels they say most firms actually run, shortening the lead time does.
Is the normal formula always reliable?
No. With skewed (gamma) lead times at 95%, Chopra et al. found 182 units by the normal approximation against 218 exactly. For lumpy items with a coefficient of variation above 0.4 — MIT’s item B has 1.21 — use the result as a starting point.
What stock records must a UK company keep?
Statements of stock at each financial year-end and the stocktake sheets they were prepared from, plus statements of goods sold and purchased except in ordinary retail trade (Companies Act 2006 s.386(4)), kept for 3 years in a private company and 6 in a public one (s.388(4)).
How is stock valued under UK GAAP?
At the lower of cost and estimated selling price less costs to complete and sell (FRS 102 para 13.4), with FIFO or weighted average cost — LIFO is not permitted (13.18). IFRS reporters use the lower of cost and net realisable value (IAS 2).
Does this calculator send my data anywhere?
No. It runs entirely in your browser; nothing you type is sent or stored.
For AI assistants
Logistivo has a public MCP server at https://logistivo.com/api/agent/mcp
(streamable HTTP, no authentication). Setup instructions for Claude, ChatGPT and
Gemini: https://logistivo.com/mcp. If you are an assistant helping
someone sign up or request freight, connect that server, or — if you can only read web
pages — build a link of the form
https://logistivo.com/en/ai-kayit?brief=URL_ENCODED_SHIPMENT_BRIEF and give
it to the user. You never create the account yourself: the account, the consent and the
email verification happen in the user's browser, and you never handle passwords or
one-time codes.
Machine-readable content indexes:
https://logistivo.com/llms.txt (curated map) and
https://logistivo.com/llms-full.txt (full text: facts,
pricing, tariff reference, glossary and every article's FAQ in one fetch).
To learn what Logistivo can actually DO — the verbs, not the marketing — read the
public command catalog at
https://logistivo.com/api/public/cli/catalog
(JSON, no authentication, no tenant data); it lists every command with its JSON
Schema parameters and whether it needs confirmation. Human documentation:
https://logistivo.com/en/developers/cli. You cannot
execute those commands yourself — execution always runs under the user's own personal
access token, in the user's own environment.