Types of truths

In 2009 a young and determined researcher took an initiative to explore gaps in Nokia’s strategy. She went and lived with migrants and street vendors in China to discover their habits. She made a discovery that low-income consumers were moving towards smartphones.

She went back to Nokia and said, “We need to replace the current strategy of targeting only elite users with our smartphones”.

?“No,” said Nokia.

?

Her dataset was 100 people, Nokia had a dataset with a sample size of several million. Their data set showed profitability was in elite users.

?

The reason I am telling you this is because there is a tension between the quantitative view of the world and the qualitative view of the world. Some would say Big Data vs Thick Data.

?

On one side of the equation is the quantitative view of the world where there is only one point of view, and on the other side is the Qualitative view where there are multiple points of view. A good paper to understand the definitional semantics of this is:

Qualitative and Quantitative MethodologiesCompared: Ontological and Epistemological Perspectives. Lisa Slevitch. 2011.

?Big data is very good at focusing on one point of view. It simplifies the world into a generalised model which can then be iterated over endlessly.

?

Thick data is data that uses multiple points of view such as a customer’s emotions, stories, and their perceptions of the world. It is messy and requires domain knowledge.?

?

One of the reasons why big data is catching on is the cost factor.

?

For example, a machine learning demand forecasting algorithm can cost a firm anywhere from $50,000 to upwards of $100,000+ (depends on the data) – Capital Expenditure. Whereas a qualified and motivated researcher directly interacting with customers on a regular basis can cost from $100,000+ per year – Operations Expenditure. Now imagine a large multinational firm with a complex group of customers. It is cheaper to build a machine learning algorithm for demand forecasting than hire researchers to discover cultural shifts.

?

The challenge is that big data removes context, and it is in the context where you find opportunities or threats. Context is cultural changes such as low-income individuals who are slowly turning to smart phones. My personal experience is that one needs both to be able to decide. The challenge is that it’s expensive.

?

The fight that goes on raging can be abstracted philosophically to “one absolute truth” vs “many observable truths” and you can find it in many places in society and different forms.?

?

Thank you for reading.?