Level 3 · Same average, different days with STDEV.S

Rachel’s two categories are a near-perfect test case. Over January 2025, Textiles made $21,519.22 and Outdoor $21,520.57: the same average day, to the dollar. A report that shows only the average says they behave alike. They don’t, and the standard deviation is the number that proves it.

What the standard deviation measures

The standard deviation is the typical distance between a value and the mean. Small means the days cluster near the average; large means they swing far above and below it. It is in the same unit as the data, here dollars per day.

=STDEV.S(B2:B32)

On this sheet:

	Textiles	Outdoor
Average	$694.17	$694.21
Median	$666.88	$72.99
Standard deviation	$495.07	$1,202.07
Std dev ÷ average	71.3%	173.2%

For Textiles, a day is typically about $495 away from $694. For Outdoor it is about $1,202 away: more than the average itself.

What the spread looks like in the rows

Scroll column C. Outdoor sold nothing at all on 13 of the 31 days, and three days (January 2, 12 and 25) brought in $11,458.40 between them, more than half the month. That is why its median is only $72.99: on a typical January day Outdoor sells almost nothing, then one big order lands. Textiles sold something every day, from $26.99 to $1,848.80.

So the equal averages describe two different businesses. A cash buffer sized on $694 a day is comfortable for Textiles and meaningless for Outdoor.

Comparing spreads: divide by the mean

A standard deviation of $500 is large for a category that averages $700 a day and tiny for one that averages $50,000. To compare spreads across series of different sizes, divide by the mean:

=F4/F2

This ratio is called the coefficient of variation. Here the means are equal, so it confirms what the raw numbers say, but it is the habit to keep: it lets you compare a small category with a big one, or January with December.

STDEV.S, STDEV.P and STDEV

Function	Divides by	Use when
STDEV.S	n − 1	the data is a sample of a bigger process (the usual case)
STDEV.P	n	the data is the entire population you care about
STDEV	n − 1	the old name for STDEV.S; same result

The next level is about choosing between them. Google Sheets has all three, plus STDEVP.

Common mistakes

• Dropping zero days. A day with no sales is a value of 0, not a missing value. Here the query fills them in; in raw exports they are often missing rows you have to add back.
• Reading the standard deviation without the mean. $1,202 means nothing until you know the average is $694.
• Including the summary cells in the range. =STDEV.S(B:B) would also pick up anything you type lower in the column.
• Locking the range with $ before filling right: $B$2:$B$32 filled to G still reads Textiles.