Quantitative sales forecasting. ** Flashcards
What are moving averages?
Moving averages are a statistical technique used in sales forecasting to identify trends and patterns in historical sales data. They calculate the average value of a set of numbers over a specified time period and are useful for identifying seasonal variations and long-term trends.
What is time series data?
When data can be collected at consistent time intervals and presented in time order.
What is time series analysis?
The analysis that is used to reveal underlying patterns in time series data.
What are the difficulties of time series analysis?
Can be difficult to interpret if the data fluctuates - unclear what the relevant patterns actually are.
What do businesses do to counter unclear time series data?
They can take moving averages of the data which smooths out the fluctuations and make patterns easier to identify - they do this usually by amalgamating data over longer periods of time, and looking at patterns based on the longer time periods.
What are trends in analysis terms?
The long term movement of a variable.
What are three period moving averages?
These calculate the averages for data from periods of 1,2 and 3 (all specific time period - usually months), then 2,3 and 4 and so on.
What does this table show? Why are there no figures for 1 and 5
It shows the annual sales of a business over a 5 year period. The three year moving averages have also been calculated, year two has been calculated as an average from years 1,2 and 3.
There are no averages for the first and last rows as they would need the sales values for before year 1 and after year 5.
The average is always placed in the middle of the time period, the 76 is the middle of 1,2 and 3.
What are four quarter moving averages?
These take the averages over the quarters from the year, this can remove seasonal variations in the data.
How are four quarter moving averages done? What is centring. (Refer to table to practice).
It’s the same as the three moving averages, only as there is no clear middle point, after working out the mean for 4 months eg (1,2,3,4) you add find the mean from the next moving average (2,3,4,5) and add the two means together and divide by 2. This is known as centring, where you find the average of two four quarter moving averages and place this against the third quarter of the first moving average.
What are scatter graphs used for?
Sales data over time can be displayed as a scatter graph to see trends such as sales of a product over a number of years.
Having collected your data, what can you do with it?
You plot it on a graph and draw a line of best fit through all points which shows the overall trends in the data.
What can a line of best fit show a business in terms of trends?
You can look for a correlation between two variables. This can show how closely two variable are related. The closer the line of best fit is to the data points, the stronger the correlation is.
How useful are scatter graphs?
They are useful, but don’t show the causes and effects. For example a strong correlation between monthly sales and the amount spent on promotion doesn’t mean the two are inextricably linked. There will probably be other factors also involved.
What is extrapolation? Why is it used? (Analyse the chart)
Trends in past sales data can be continued into the future (extrapolated) to forecast future sales, which enables managers to set targets. Sales performance can also be measured against these targets.
Can you extrapolate on a graph without a line of best fit?
Yes. If the graph has a curve and you have data for the recent trends, you can guess for the future. `
Why may extrapolating a graph without a line of best fit be inaccurate?
This method mostly focusses on recent changes, but it is hard to know if these changes will continue or whether considering long term trends are more important, which may indicate things will go back to how they were.
Are extrapolations consistently accurate?
No; they rely mostly on past data and trends repeating themselves, however this is never guaranteed, which can make extrapolations inaccurate.
Where are extrapolations most suitable?
In more stable markets with little external opportunities and threats.
Where would extrapolation be applied in dynamic markets?
For predicting short term steps, such as 3 quarters ahead.
Why will businesses have to account for variations in trends?
Because businesses are often affected by cycles that change sales data, such as seasons or product life cycles which cause regular changes in sales.
How may a business extrapolate using data, instead of lines of best fit?
Through finding the cyclical and average variations in their sales data.
Formula for cyclical variation?
Actual sales figures - moving average
What are the uses of cyclical variations?
They highlight the extent to which specific periods cause variations in their sales, we can see if specific time periods are going to be higher than their moving averages to get more reliable data. By looking at this, a business can ascertain when they are likely to see surges in demand and drops in demand which can help them with decision making.
What is the formula for average cyclical variation
Total cyclical variations/ number of cyclical variations
