Black Friday recurs every year on the Friday after the fourth Thursday in November.
Its seasonal influence is predictable, but it does not occupy the same calendar date—or always the same 52-week lag.
Events & holidays: concepts and Northstar examples. Illustrative Northstar examples, not measured business results.
Concept
What it means
Northstar example
📍Align by event, not just by date.
Create a Black Friday indicator and event-relative windows: the lead-up, Friday, weekend, and Cyber Monday.
This separates the recurring event from ordinary weekday and annual effects.
🛒 Represent Northstar's Black Friday lead-up, Friday, weekend, and Cyber Monday by their distance from the actual event date.
🔮The event date is known; the lift isn’t.
Future calendar indicators are available at forecast time.
Discount depth, advertising, stock, and consumer behavior affect the size of the lift.
If those inputs are unknown, use scenarios.
🛒 Northstar knows next year's Black Friday date now, but not the resulting sales.
Discount depth and ad spending may need explicit scenarios.
🧪Don’t confuse promotion with causality.
A regression coefficient is an association conditional on the model.
Promotions may be scheduled when demand would already be high, making causal interpretation difficult.
🛒 A large holiday-promotion coefficient does not prove the promotion caused all the extra sales; Northstar may have scheduled it during an already busy season.
FIG 17.1
Follow the event, not the date
WHAT THIS SHOWS
Align Northstar's Black Friday sales by the event date or the ordinary calendar date and compare where the peaks land.
The question it answers: Why can a fixed-date annual slot misalign a moving holiday?
Align sales by
2023 · Nov 242024 · Nov 292025 · Nov 282026 · Nov 27
2023Nov 24
2024Nov 29
2025Nov 28
2026Nov 27
These years use the same illustrative event lift. Event alignment reveals that identical shape; calendar-date alignment shifts it. In real data, discounting, stock, and demand also change the peak height.
Fine print & assumptions
Constructed teaching data and illustrative calculations, not a production forecasting service. Read the diagram's method notes and distinguish supplied simulator components from fitted estimates.
CHECK YOURSELF · CHOOSE AN ANSWER
Which feature can safely be known before next Black Friday?