Quantitative Sales Forecasting Flashcards
What’s time-series data
Sales figures being collected at consistent time intervals (e.g. every month) and presented in time order
Why are moving averages taken
Because time-series data can be difficult to interpret if the data has fluctuations
- Moving averages smooths out the fluctuations in the data, making it easier to identify trends
How to work out a 3 year moving average of year 2
Add year 1, 2 and 3 together then divide it by 3
How to work out a 4 year moving average, placing it against the 3rd quarter of the moving average
Add year 1, 2, 3 and 4 together then divide it by 4
Then add year 2, 3, 4 and 5 together then divide it by 4
Add the 2 answers together and divide by 2, which is the answer placed next to year 3
What graphs show trends in data (used for moving averages to spot trends) in sales
Scatter graph, with use of a line of best fit
What’s extrapolation used for
Used to predict future sales, as line of best fit continues beyond the points
What’s a disadvantage of extrapolation
Relies on the assumption that past trends will remain true
- sales performance can be heavily influenced by external (change in consumer preference, technology) and internal factors (price, quality change) Therefore extrapolations don’t predict future sales very accurately
When’s extrapolation most useful
Fairly stable environments
E.g. size of market or number of competitors is unlikely to change
- In dynamic markets it’s best to use extrapolation for predicting just a few time steps (e.g. 3 quarters) ahead as future is more uncertain
The difference between actual sales figures and moving average for a period is called what
Cyclical variation
= Sales figures - moving average
Whys cyclical variations used
Because actual sales can differ from the moving average depending on where in its cycle the business is
Limitations of quantitative sales forecasts
- Forecast is unreliable if it forecasts a long time in the future
- If the market is fast changing, then wont be accurate
- Those preparing the forecast may not have a good understanding of how to use data to produce a forecast