Data collection methods have undergone a revolution in recent times. Prior to the nineties, market surveys served as a ‘single’ source for data collection on product preferences and demand for the product at various prices. Based on this, the firms would then decide on prices. The survey data are structured to answer some particular questions and therefore are called ‘designed’ or ‘structured’ data.
The collection method is intrusive, costly, and limited, so no wonder that response rates to these surveys have been declining over the years. Now standing against it, are the modern methods of data collection which involve collecting data from multiple sources. The search engines of e-business firms use Machine Learning (ML), deep learning algorithms, store the data on individual behaviours of customers, analyse their consumption patterns and also provide recommendations to customers (nudge) based on this data. The data so collected are mostly non-intrusive, continuous, dense, and organic. In brief, with every click of a mouse (pad), the information trail is captured and stored in the digital space. In addition to that, social platforms such as, Facebook, Instagram, etc., are also used to collect the data on consumer behaviour and their tastes, etc. It may also be pointed out at this juncture that big data also are complemented with the small survey data in some instances, especially in spheres of public policy.
The analysis of the big data using data science provides ‘insights’ to companies. Usually, the firms, depending on their stage of development, set their objectives either as ‘profit’ or ‘sales’ maximization, which involves three major decisions: (i) quantities to be produced and sold of various products/services (product mix); (ii) pricing strategy; and (iii) cost reductions.
Data science and sales maximization: With the help of data science, the firms try to increase ‘revenue’ through measures such as retaining old customers, adding new customers through referrals of old ones, recommending products to said customers, deciding on optimum product mix, and through advertising. This becomes possible with the help of big data captured by companies when a consumer searches for a product, buys the product, and shares info (positive or negative) about it through social media or any other platform. Profitability of a company can also be increased by cross-selling the add-on products associated with the main one, or nudging the consumer to buy the latest version for some additional money.
Data science algorithms are able to decipher the various attributes and thus correlate among products and match them to the tastes of the consumers. Very often, a consumer would find that a search for a product yields recommendation for a similar one. The behaviour of the consumer, irrespective of whether they buy it or don’t, provides data history on purchases to the company in a non-intrusive manner which can be used for marketing strategy. The companies often use the ‘recommendation engines’ for prediction purposes. Creation of customer profiles, including their location, the type of products they are interested in, the kind of nudging that is effective, etc., has become possible with integration of data sources from the purchase history of the consumers and social media. The task of the product team is to identify the optimum product mix, and the quantities of it that should be supplied at various points of time. Data science enables ecommerce businesses with both advanced predictive and prescriptive models.
Data science and pricing strategies: Data science uses algorithms and Machine Learning (ML) for ‘dynamic pricing’, i.e., setting optimal prices. Usually, pricing algorithms combine the use of (Artificial Intelligence) AI and ML, which enables setting the prices for different target groups. This is done depending on market trends, demand fluctuations, consumer behaviour, purchasing power, and a host of other factors. The dynamic pricing algorithms also consider the relationship between the price and the quantity demanded, and it can be observed that the periods of high demand are associated with higher prices. Thus, prices can vary over time on the same day, depending on algorithms used, which can thus enable the firms to maximize their profits. This can also work in reverse and nudge the consumers to look out for times when the demand is not so high to reap the benefits of lower prices.
Data science and cost reductions: By making use of big data and data science, firms also try to reduce costs of production though other various methods. On one hand, data science enables companies to predict efficient supply-chain models for speedy and cost-efficient product delivery, provides data about location of consumers, producers and transportation costs, warehouses, inventories, etc, while on the other hand, it is used for modelling purposes to predict efficient supply chains and delivering the appropriate products to appropriate locations at appropriate times. This analysis and prediction reduce costs which yet again maximises profits.
To sum up, data science has revolutionised the way information about consumers can be collected, processed, and analysed. The decisions about product-mix and price costs can be made depending on various variables in a dynamic manner, and most importantly, all this can be achieved in a non-intrusive manner for different target groups using different nudging strategies.
Views expressed above are the author’s own.
END OF ARTICLE