Reference no: EM132286123
Nestle has a variety of big data analysts in their headquarters back in Vevey, Switzerland. One enterprising analyst, comparing sales performance of various product lines between different types of countries, notices what appears to be a recurring anomaly.
More precisely, he discovers that even though the GDP per capita report for a country might be low to extremely low, grassroots data on actual sales levels and trends in those markets, such as in Mozambique, Guatemala, Zambia, Cambodia, indicate surprisingly robust demand for the sorts of food products that Nestlé makes. In trying to reduce the variance, the analyst runs and reruns the data, but cannot resolve the problem.
In frustration, he consults a colleague, asking what she thinks might be going on with his data. She immediately concurs with his outlook—that is, if we were to stick with the officially reported GDP per capita data, Nestlé would wisely would postpone expanding into many of these countries. Alternatively, she points out, if we rely upon the actual sales data reported through various sources, then Nestlé should expand operations these markets as soon as possible.
Brainstorming some more, the analysts come up with a solution that they believe will more accurately estimate current sales levels and potential growth in many of these markets.
Given your understanding of economic dynamics in many developing markets, such as Mozambique, Cambodia, and the like, which of the following assumptions Do you believe the analysts relied upon in developing their solution.
A In some markets, particularly in developing countries, the shadow economy, whereby transactions are in cash in evade official measurement, can be quite extensive.
B People are intrinsically unpredictable and inexplicable anomalies are inevitable.
C Consuming behavior in countries with low to extremely low GDP per capita are difficult to model.
D Aggregating data across all markets, both developed and developing, and then smoothing the variability through logarithmic transformations will likely improve the accuracy of the model.