# Quantitative sales forecasting

?
• Created by: noe
• Created on: 18-09-20 09:22
• Quantitative sales forecasting
• Time-series analysis: method that allows a business to predict future levels from past figures
• Main components a business wants to identify in time-series analysis: trend and seasonal, cyclical and random fluctuations.
• Accurate sales forecasts helps businesses to make a number of key decisions. E.g. how much stock to hold or how many people to employ.
• Identifying the trend: analysis of figures will tell a business whether there is an upward, downward or constant trend.
• Moving average: a succession of averages derived from successive segments of a series of values
• When plotted onto a graph, line is 'smoother' removing any fluctiations in sales each year + gives a more obvious picture of the trend.
• Three-year moving average
• Add up sales for the first 3 years and divide by 3 Then, the first year's sales dropped out and the next year's sales added to calculate average, and so on.
• Centring: method used to calculate a moving average, where the average is plotted or calculated in relation to the central figure.
• Four-year moving average: Add up first four years, drop out the first year and add the fifth, add those up and add the answer to the first four years' average and divide by 8
• Extrapolation: use of past sales data to forecast future sales.
• Sales figures can be predicted by drawing a line of best fit through the trend figures and extending it making sure it matches the general slopes of all points in the trend.
• However, prediction is made believing everything will stay the same but, if other factors changed, it would be inaccurate.
• Seasonal variations. E.g. a three-year period divided in quarters.
• Variations from the trend
• Actual sales - trend. If the answer is positive it is said to be favourable but if it's negative it would be adverse.
• Average of variations calculated by adding the variations of a certain years and diving it into the number of variations.
• Then, this would be added to the prediction value based on the trend to make it more accurate