Building upon my previous research on ‘The Rise of Big Data’, I can highlight the existence of a two-tier system of micro- and large-scale (big) data analytics. All companies that dominate local or neighbouring markets collect data at a micro-scale level for the purpose of predictive analytics, such as linear, including decision trees, vector, and cluster modelling. They use geo-demographic variables, such as income, age, and so on, and behavioural data to predict the target’s willingness to buy a particular product. Major retailers attempt to test their customers’ willingness to make certain purchases. Regularly, credit-rating companies use sample modelling to test the probability of fraud; insurance companies for the probability of claims; life insurance companies to estimate life expectancy; banks for the probability of a mortgage’s voluntary foreclosure; and so on. However, given the limited size of the sample, such predictive analytics may not prove accurate. In contrast, large-scale corporations that possess or harvest a large amount of big data may inter alia use raw data from the unstructured content of emails or the web for data mining purposes; machine-generated data; statistical software packages, such as IBM, Stata, Rapid Minder, Google’s open source software, Apache Hadoop, Revolution, and so on; and automated data that is a mix of data-driven and expert-derived rules to analyse big data. Software packages act as intelligent agents that allow for quick automation and processing of big data analytics. With the help of the quantitative and statistical analysis of big data, it is, however, possible to accurately measure the consumers’ willingness to pay for particular products, determine the elasticity of demand in response to price changes, observe trends in the life cycle of a product, identify under-performing products, and categorise customers. While the micro-scale behavioural modelling of data serves for the analysis and prediction of the risks associated with the use of targeted advertising and promotional campaigns, when the same modelling is being applied at a large-scale level to forecast customers’ demand, to predict product trends, and to make strategic pricing recommendations, the latter inevitably becomes part of a wider social experiment of intensive platform monitoring and data sharing with data analytics companies. Due to the size of the sample of participants due to be observed, the latter forecasts tend to be even more accurate and to reliably inform producers of estimated demand and future pricing options. I would argue that no marketing research harms consumers as long as the sample of the targeted consumers remains meaningful, but limited for a specific purpose. Otherwise, big data analytics is a perfect substitute for direct or indirect exchanges of strategic information regarding actual or future pricing methods; estimated demand; consumers’ preferences, location, investment; and so much more. However, larger companies or corporations are in a stronger position to extract strategic data that can later be exploited tactically, i.e., through targeted advertising, and strategically, by informing the price setting mechanism. Instead of a business-to-business exchange of information (B2B: ‘hub and spoke’ conspiracy), this large-scale marketing experiment moves on to the prospective consumers (B2C: track-and-monitor conspiracy). In my opinion, this phenomenon, which I have previously identified as a track-and-monitor conspiracy on the basis of consumers’ geographical location; socio-economic demographics, i.e., income status; and behavioural data, i.e., preferences and interests, allows for a pricing conspiracy to be implemented with the help of consumers rather than competitors. For example, consumers identified as living in remote areas, i.e., the Highlands or small islands, usually have less choice and can therefore be charged more for other terms and conditions, such as transportation costs. Knowing consumers’ category of income, businesses can more accurately predict their reservation price in terms of bargaining. It is similar to a meeting of minds between the buyer and the seller, where the latter knows how much the former is able to potentially spend.
Chirita, A. D. (2018). Written Evidence to HM Treasury's Digital Competition Expert Panel. [No known commissioning body]