3. Spread, Correlation, and Forecasting

• STDEV.S shows how far values typically stray from the average. The bigger it is, the less stable the sales;
• CORREL(x, y) measures how strongly two series are related, from -1 to 1. Close to 1 means they rise together;
• FORECAST.LINEAR(x, known_y's, known_x's) predicts a value along a straight trend line.

Correlation doesn't prove a cause, but it hints at where to look for one.