What it means
An order planned for 14 days might actually take 19, and the order-level difference is plus 5 days if the calculation is actual minus planned, which is a schedule miss, not statistical variance. Teams often call that difference a lead time variance, and plain-language use is common, but it can cause confusion when a data analyst uses the formal statistical definition, so label the column "actual minus planned days" when that is what it contains.
A positive result under this convention is late and a negative result means the order arrived sooner than planned, though some systems reverse the subtraction, so a reader needs the sign convention. The start and finish of each clock must match, since comparing dispatch-to-arrival actual time with purchase-order-to-stock expected time makes a meaningless gap, so record whether receiving and quality release are included.
A promised date is not always the planning assumption, because the supplier may promise delivery in 7 days while a cautious buyer plans around 10, so report deviation from each reference separately. An average gap can also hide inconsistency, since early and late orders can cancel each other numerically, so show the distribution of misses and the percentage beyond an operational threshold.
Statistical variance has another meaning: Penn State defines it as the average of squared differences from the mean, with a distinction between population and sample calculations, and if lead time is in days, variance is in days squared. Standard deviation is the square root of statistical variance and returns to days, so it is easier for a manager to interpret.
Lead time variability is the broader issue of inconsistent elapsed times, and statistical variance is one measure of it. A supplier may always deliver in 12 days when the plan says 10, so the deviation from plan is plus 2 days every time but the standard deviation of actual lead time is zero, which points to a planning bias rather than erratic service.
Conversely, a supplier could alternate between 8 and 12 days with a plan of 10, so the average deviation is zero yet actual times are variable and some orders are late. For a sample of observed lead times, calculate a mean, square each deviation from that mean, sum the squares and divide by one less than the number of observations, which gives sample variance rather than a per-order delivery miss; use population variance when observations constitute the full population being described.
With lead times of 8, 10 and 12 days the mean is 10 days and population variance is (4 + 0 + 4) divided by 3, or about 2.67 days squared. The sample-variance denominator for those three observations would be 2, giving 4 days squared, and the difference is a statistical choice, not a change in actual delivery times, so state which calculation the report uses.
For a supplier scorecard, show promised and actual times, signed deviations, median delay and percentage delivered within agreed tolerance, and add standard deviation for a view of consistency, because a single headline is rarely enough. Look for differences by route, product and order size, since a combined average can hide one delayed lane, and follow up a repeated late gap with a process timeline, as orders may wait for approval, supplier production, customs or warehouse intake.
ASCM notes that lead-time variability can affect safety-stock calculations, but an individual five-day miss does not automatically imply a particular safety-stock increase, and lead time variance becomes useful only with its definition attached.
In practice
Real-world examples.
Example
An order expected in 14 days arrives in 19; actual minus expected is plus 5 days. The buyer records the signed gap in a column labelled with its sign convention.
Example
A supplier consistently takes 12 days rather than a promised 10; the promise is biased, though delivery may be stable. The buyer updates the planning assumption instead of adding safety stock.
Example
A buyer records both the late-order percentage and the standard deviation across comparable orders. The two figures together show how often orders are late and how unpredictable they are.
Formula
Calculation
Order-level deviation in days = actual lead time - planned lead time. Statistical population variance = sum of (each actual lead time - mean actual lead time) squared / number of observations. Do not label one formula as the other.
Worked example. An invented buyer plans every order at 14 days, and four comparable orders actually take 19, 12, 14 and 17 days.
- Order-level deviations (actual minus planned) = +5, -2, 0 and +3 days. Their average is (5 - 2 + 0 + 3) / 4 = +1.5 days, so on average orders run late.
- Mean actual lead time = (19 + 12 + 14 + 17) / 4 = 15.5 days.
- Deviations from that mean = +3.5, -3.5, -1.5 and +1.5; squared they are 12.25, 12.25, 2.25 and 2.25, which sum to 29.
- Population variance = 29 / 4 = 7.25 days squared, and standard deviation = about 2.69 days.
The +1.5 days measures the planning bias, while the 2.69 days measures how unpredictable the orders are, and the two answer different questions.Case study
Seen in the real world.
In this entirely fictional case, Seabright Components records a plus-five-day order-level deviation for one delivery. Its buyer then checks twenty comparable orders rather than changing its whole inventory policy from one incident. The supplier proves consistently slower than the old planning assumption.
The buyer revises the assumption and separately tracks whether the actual times become more predictable. The buyer also splits the twenty orders by route and finds in this invented story that one lane accounts for most of the late gap. The follow-up conversation with the supplier therefore focuses on that lane, not on every delivery.
Watch out
Common mistakes.
- Using "variance" without defining plan deviation or statistical variance.
- Letting early and late deviations cancel in an average.
- Comparing different clock boundaries or day-count rules.
Questions
People also ask.
Does plus five mean five days late?
Yes, if the report defines its signed gap as actual minus planned lead time.
What unit does statistical variance use?
Squared time units, such as days squared; standard deviation is expressed in days.
Can a stable supplier miss every promise?
Yes. Predictability and accuracy against a promised time are separate measures.
From the founder's library
Take it further with the book.
Build your financial confidence beyond this definition. Shihan's full-length guide, Accounting Fundamentals, takes the same plain-English approach and turns it into a complete, practical playbook for non-finance managers, business owners and students - with chapter-end quiz answers and presentation slides included.
25% off with code MMHQ25, applied at checkout. Priced in USD - checkout may show the equivalent in your local currency.
View the book and save 25%