Sales forecasting and time-series
Sales forecasting
- Marketing analysis uses data to make better decisions.
- Sales forecasting predicts future sales from past data — helps plan production, staff, stock and cash.
Time-series analysis
- Time-series analysis finds a pattern in past sales over time:
- the trend — the long-run direction (up/down/flat),
- the seasonal variation — the regular within-year rise and fall.
- A moving average smooths the wobbles to reveal the trend.
$$\text{seasonal variation} = \text{actual value} - \text{trend value}$$
- A forecast extrapolates the trend, then adds the usual seasonal variation (only valid if patterns continue).
Practice
In time-series analysis, the trend is:
The trend is the long-run direction; seasonal variation is the regular within-year pattern.
Practice
Actual sales are 520 and the trend value is 500. What is the seasonal variation?
Seasonal variation = actual − trend = 520 − 500 = 20.
Practice
Extrapolation extends the past trend into the future and assumes patterns continue.
Forecasts extrapolate the trend; they are only reliable if past patterns keep holding.
You've got it
Key idea
- sales forecasting plans ahead from past data
- time-series = trend + seasonal variation; a moving average smooths to find the trend
- seasonal variation = actual − trend; forecasting extrapolates the trend