3.3.1 - Quantitative Sales Forecasting

Spotting the trend

When a business looks at sales over time, the raw figures often jump around because of promotions, weather, one-off events or short-term changes in consumer behaviour. Time-series analysis tries to smooth some of that noise so the underlying trend becomes easier to see. If managers can see the real direction of sales more clearly, they can make better planning decisions and therefore reduce the risk of overstocking or underestimating demand.

One of the main quantitative methods is the moving average. A moving average takes a set number of past periods, finds the average, then moves forward one period and repeats the process. Because each average combines several observations, unusual highs and lows have less influence, so the business gets a clearer view of the trend.

A three-period moving average uses three consecutive time periods and is placed on the middle period. That placement matters in exams: the average of Years 1, 2 and 3 belongs on Year 2, not Year 3.

Here is a simple worked example for annual sales revenue:

The Year 2 moving average is (18 + 21 + 21) / 3 = 20. The Year 3 moving average is (21 + 21 + 24) / 3 = 22. As the moving averages rise, the business can see an upward trend even though the raw data does not increase by the same amount every year. That matters because a clearer trend makes future planning more accurate.

Four-quarter averages and centring

Quarterly data often contains strong seasonal patterns, so a business may use a four-quarter moving average instead of a three-period one. This is still the same idea of smoothing data, but it uses four consecutive quarters rather than three yearly periods.

The key difference is that four is an even number. That means the first four-quarter average sits between two quarters rather than directly on one quarter. To attach the trend to an actual quarter, the business centres the data by averaging two adjacent four-quarter moving averages.

A clean exam method is shown below:

The reason for centring is simple: without it, the average is not aligned to a real quarter, so the trend line would be misplaced. Once centred, managers can compare the trend with actual quarterly sales more accurately and distinguish underlying growth from seasonal movements. A smoothed trend alone does not reproduce each seasonal peak: staffing and stock plans must also consider the seasonal pattern.

Exam anchor: a three-period moving average sits on the middle period, but a four-quarter moving average usually has to be centred first.

Scatter graphs, line of best fit and extrapolation

A business can also forecast sales by comparing sales with another variable, such as income, temperature, or advertising spend. A scatter graph plots the two variables together. If the points show a clear pattern, managers can draw a line of best fit through the middle of the data to summarise the relationship.