This compares weekly sales across the three store types, where the box shows where most weeks land. Type A stores sell the most, Type B sits in the middle, and Type C is the smallest with sales packed in a tight range. 

This grid shows how closely each factor moves with weekly sales, where closer to 1 means they rise together. Store size stands out at 0.81, far ahead of everything else. Discount spend is weak at 0.23, and temperature, fuel price, and unemployment barely register.

This shows how much each factor pushes weekly sales up (green) or down (red) when everything else is held steady. Holidays add roughly $46K a week, while higher fuel prices and unemployment pull sales down. Size and discount bars look tiny only because they’re measured per square foot and per dollar, and store type is best read as a direction, not an exact amount.

Each dot is one store week, plotted by what really happened against what the model predicted. The closer to the dashed line the better, and most dots hug it, which is why the model scores 0.93.

This shows the model’s predicted sales as discount spend rises from $0 to about $32K. The line climbs, but check the scale on the left: the whole move is only about $5K on a $1.06M week, roughly half a percent.