The growth of “big data” changes the very essence of modern markets in an important sense. Big data are nothing but the digital traces of a growing number of people’s daily transactions, activities and movements, which are automatically recorded by digital devices and end up in huge amounts in the hands of companies and governments. Payments by debit and credit cards record timing, place, amount, and identity of payer and payee; supermarket loyalty cards report purchases by type, quantity, price, date; frequent traveler programs and public transport cards log users’ locations and movements; and CCTV cameras in retail centers, buses and urban streets capture details from clothing and gestures to facial expressions.
This means that all our market transactions – purchases and sales – are identifiable, and our card providers know a great deal about our economic actions. Our consumption habits (and income and tastes) may seem more opaque to scrutiny but at least to some extent, can be inferred from our locations, movements, and detail of expenses. If I buy some beer, maybe my supermarket cannot tell much about my drinking; but if I never buy any alcohol, it will have strong reasons to conclude that I am unlikely to get drunk. As data crunching techniques progress (admittedly, they are still in their infancy now), my supermarket will get better and better at gauging my habits, practices and preferences.
Systematic identification and trackability change the view held in economics from Adam Smith onward, of markets characterized by mass production, standardization of goods and services, and use of equal-for-all paper money as a universal means of exchange. In this world, a buyer and a seller could settle a trade with hardly any need for personal identification. For classical economics, anonymity was new and they resolutely opposed it to the personal net of mutual favors, gifts and family solidarities that was typical of pre-modern, feudal communities. Anonymity appeared as a major defining characteristic of modern markets.
Granted, the traditional business concept of “segmentation” already distinguished customers into groups by some defining characteristic: For example in air or rail travel, wealth distinguished first class from economy class passengers. Yet detailed knowledge of each passenger’s circumstances was hardly available and typically, a company relied only on few aggregate data about total demand for the two classes. The trick was to leave to customers the task of selecting themselves into the most appropriate class: the company should just post two different prices for a journey and make the difference strong enough to incite richer customers to buy the higher fare. No further personal identification needed. Jules Dupuit made this principle explicit as early as 1844, when he was reasoning on ticket pricing for the (then newly-created) French state rail company.
We are now entering a phase (can we possibly call it post-modern?) in which the introduction of cards, loyalty schemes, smartphones and other mobile digital devices is bringing back identification as in pre-modern times, but on a much larger scale. This changes our view of markets: instead of producing for the masses, companies can now hope to offer much more finely tailored and targeted products and services. In an influential report on Big Data, McKinsey says that the traditional business concept of ‘segmentation’ is gradually being replaced with ‘personalization’. Use of increasingly granular data in almost-real time opens the way to practicing price discrimination on a much larger scale than before, so as to extract value from all possible customer types and market niches. So instead of offering just two classes of travel, airlines and railways can now offer distinct packages for each particular passenger, based on their own unique characteristics and preferences. Obviously, more passengers would find a suitable travel solution, and would buy tickets; in economics’ terminology, both consumer and producer surplus would be maximized.
Unsurprisingly, the corporate world has expressed great interest for the potential of digital data to enhance, among other things, marketing, customer relationships and sales management. It is often claimed that potential benefits will accrue to both firms and final consumers: for example, McKinsey estimates that services enabled by personal-location data alone, can enable consumers to capture $600 billion in economic surplus.
All this is not yet reality, and the privacy protection issues that Big Data are raising still need to be addressed before any prediction can be made. But the current transformations might substantially modify the functioning of the market mechanism, as different people would systematically pay different prices for the same good. There is research to be done on what this would imply economically.