Statistical
SLOPE function: estimate variable cost per unit or trend rate of change
Calculate ordinary least-squares linear slope (change in Y per unit of X) using SLOPE—without reversing the Y and X arguments.
Put known_y's FIRST and known_x's SECOND
The #1 mistake with SLOPE(known_y's, known_x's) is passing X before Y. Because slope measures ΔY / ΔX (dollars per unit, revenue per month), the dependent outcome column (TotalCost in B2:B5) goes first and the driver column (Units in A2:A5) goes second.
Reversing the arguments fits X against Y instead. In this perfectly linear example the result is the reciprocal, but that reciprocal relationship does not generally hold for noisy observations.
| UnitsProduced (A) | TotalCost (B) |
|---|---|
| 10 | 650 |
| 20 | 800 |
| 30 | 950 |
| 40 | 1100 |
Extract marginal cost per unit
Each 10-unit step adds $150 to TotalCost, so SLOPE(B2:B5, A2:A5) returns exactly $15 per unit.
=SLOPE(B2:B5, A2:A5)Zero X variance returns #DIV/0!
If every X value is identical, ΔX variance is zero (a vertical line) and SLOPE returns #DIV/0!.
=SLOPE(B2:B5, {10;10;10;10})Try the example
Use the original values or change them to test your understanding.