LOADING

Define Mobile Menu

Correlation is the relationship between two variables. Extrapolation is related to correlation, as it involves plotting a linear regression curve (also known as a line of best fit) which is based upon how two variables are correlated, and extending the line to predict future values. For Nintendo, a computer game and console manufacturing company, a common example would be the relationship between the level of advertising expenditure and the effect of this on sales of its products.

The graph to the left shows a strong positive correlation between sales and advertising, meaning if Nintendo increases its advertising, they can expect to see an increase in sales. Because all values are close to the line, as advertising expenditure increases on the x-axis, sales on the y-axis increase proportionately. The regression line can be extended past the ’80’ figure (meaning �80million) to predict what sales will be like if they spend �100million or even �900million.

The uses of such analysis to Nintendo are to an extent very valuable. By creating a scatter graph like the one above, Nintendo can spot trends and forecast future events. For example, they could estimate demand at a certain point in the coming year. One would expect demand to surge at Christmas time when people are buying each other gifts, and dip in January when (generally) consumers have less money due to spending so much on presents, and figures could be devised to estimate how many units will need to be produced.

This allows Nintendo’s operations department to build up stocks of goods beforehand in preparation, for example the ‘Nintendo Wii’ which was under stocked in all major retailers, and hence was selling on eBay (the online auctioning site) for triple the price. Because games consoles are durable goods and do not perish, Nintendo could store them in a warehouse, so the manufacturing facilities aren’t exhausted around November time. This would be positive to both Nintendo who can sell a lot of units and thus increase profit, and consumers who will not be left wanting a Wii.

In addition to this, it could aid Nintendo when designing a new advertising campaign. They could gather past data from previous campaigns and plot the values on a scatter graph to see if there is any correlation. This would aid decision making for the management staff – as they could see whether the operation would be profitable on the whole, and ultimately be worth doing. Albeit, this would only be an estimate of the intended benefit, and in reality things are much more complicated.

Following on from the previous paragraph, the main downfall of this analysis is the fact it assumes that previous patterns will continue into the future, and doesn’t account for qualitative factors. Qualitative means customers opinions and is not numerically measured. This is a huge factor when considering extrapolating for Nintendo, as technological advancements could revitalise the way games consoles are played in this rapidly changing market. A prime example is the growth of online gaming, which Nintendo have not responded to as quickly as its main competitors the x-box and the PS3. Many years ago when online gaming was not as popular, this would have had a minimal impact on sales, but the data today is less reliable and can make predictions redundant if they don’t account for such factors and changes in consumer tastes.

A final disadvantage is the effect of competitors on the validity of the data used to forecast values. Nintendo operates in an oligopolistic market (a market where there are a few large firms) where there tends to be less price competition. The other major players in the market are Microsoft and Sony with the x-box and PS3 products respectively. Said businesses have to take into account likely reactions of rivals to any change in price and output, which can be unpredictable and therefore damage the precision of statistical analysis. Nintendo could predict a �200billion increase in sales by running a bundle promotion, but the forecast would be nullified in Sony ran an equal campaign, which ultimate leads to Nintendo losing a large amount of money.

The extent to which correlation and extrapolation are useful for Nintendo for forecasting sales is arguable. On one side, it seems foolish not to forecast thoroughly before planning on running a multi-million pound operation, but on the other, in the ever changing world of business (and especially the computer games industry) the validity and accuracy of the forecasts could be destroyed in a few months with a new technological breakthrough from one of Nintendo’s competitors.

It also depends on how strong the correlation is between the two variables. A weak correlation is likely to have a higher degree of uncertainty than a relationship that is strong. It is up to the businesses marketing team whether to go project values based upon a weak correlation. Outliers or (anomalous results) can also hinder the accuracy of the regression line unless they are ignored, which again is a decision of Nintendo. The level of uncertainty makes me conclude that ultimately such analytical tools are not of use to Nintendo.

Ultimately the use depends on all the previously mentioned factors, some which cannot be accounted for. This leads me to take the conclusion that these statistical tools are not of great worth to Nintendo and they should use a fixed point moving average method to smooth out the overall and underlying trends in their forecasts. This will reduce uncertainty and remove some of the disadvantages and problems which is more beneficial for a large firm such as Nintendo.