Converting Data to Information
Marta Gonzaléz has been considering setting up a new business. One of her good friends in Tenerife, Consuela Martín, has been operating a café for a few years and she is interested in finding out the cost structure of the business. As Marta is thinking of establishing her business in the Costa Blanca, Consuela sees no threat from Marta‟s proposed operation and provides her with the following information.
From this table, Marta can see that the cost of a cortado is roughly 75 cents (1,520/2,000 for example) but this does not really give her any indication of the cost structure of the business. So, she decides to plot the data and see how it looks. This is shown in the following graph.
Two things are immediately apparent to Marta. First, the costs increase in a reasonably straight line. This makes intuitive sense, as the more drinks served means more labour, coffee beans and milk. What the graph does not tell her, however, is how the costs are split been fixed and variable costs. To get an idea of the type of costs involved, she undertakes an exercise in linear regression, a technique she remembers from her undergraduate days. The main purpose of this technique is to fit the best straight line through the data so that she can see how the costs are structured.
Although regression analysis has many parts, the most important part of the output is the equation of the line through the points in graph 1. The equation for total costs, as given in the output is:
Total Costs = €1,123 + €0.21 * (cortados sold)
So, from this equation, we can see that €1,123 are the average monthly fixed costs of Consuela‟s business and each time a cortado is served, total costs increase by €0.21. It is interesting to note at this point that around 70% of Consuela‟s costs are fixed (€1,100 out of a monthly average of around €1,600) and the remainder variable. She is a little concerned about this, but she knows that, based on Consuela‟s figures, she needs to budget €13,500 (€1,123 per month for 12 months) in the first year of operation just to keep the business afloat.
Using the equation, Marta checks how well the model performs. For the January figures, for example, she multiplies the number of cortados (2,000) by €0.21 and adds on the modelled fixed costs of €1,123. This gives a total of €1,542. This figure is very close to the actual figure of €1,520 with an error of only 1½%. In fact, this is the largest error of the 12 months. The full results are shown in table 2.
After judging these results, Marta has decided to investigate opening a coffee bar more thoroughly, as we shall see in later papers in the series. The business will be called “Cortados „R‟ Us”.