This piece examines the intersection of technology and economics, an intersection that has regularly occurred in myriads of ways. My own career as an economics student in graduate school included several inquiries into the economics of technological change, including my 1973 dissertation that built a model explaining the process of technological innovation in post-WWII electric power generation. [Pretty exciting, yes?] This topic turned out to be fortuitous once OPEC initiated its thoroughly-disruptive "Arab oil embargo" on the US in October 1973 and caused anyone who already had some knowledge of the suddenly-important energy sector to be in great demand. Lucky me.
Enough history. Technological change and advancement always have economic consequences, some immediate, some longer-term. We are still experiencing long-lasting consequences of the 19th century Industrial Revolution. There are winners and losers as well as transformations in economic relationships due to technological change. Like virtually all economic actions, many of the consequences will be unforeseen and unintended. This blog considers the technological emergence of BIG Data.
Over the past year or so the phrase BIG Data has been popularized and discussed in many forums, including Dilbert. BIG Data refers to the collection and analysis of increasing
Six months ago, the New York Times published an article, "The Age of Big Data," by Steve Lohr, a technology reporter for the Times who has followed BIG Data. Mr. Lohr talked about the impressive growth of BIG Data and the resulting additional need for analysts, programmers and managers who are conversant with data-driven, "deep analysis" methods in fields spanning artificial intelligence, communications, manufacturing, sports (e.g., Moneyball), law enforcement, shipping and retail sales. Estimates of such staffing needs run as high as one million persons. As a long-confirmed empiricist, I am mostly heartened that data, and analysis of it, are now gaining access to the drivers-seat.
Everyone, including you and me, who is connected to and/or communicates via digital devices, who has a FaceBook, LinkedIn and Twitter account has made contributions to BIG Data. And it's not just by using your mobile phone or computer. Every time you swipe your Safeway, CVS or Target card, you have volunteered more information into BIG Data. Do you benefit from providing this information? It seems like it; the last time I went to the grocery store, I was notified on the receipt that I saved $7.19 by being a member of the store's customer "club." In addition, I also benefit by having some reassurance that the store will continue to stock the items I keep purchasing. [Unfortunately, there is no such thing as 100% reassurance. One grocery store we shop at seems not to get our purchasing messages. It has repeatedly stopped carrying unique items I considered essential – like the store's 5-star worthy brand of corn chips. So it goes with my continuing commercial inputs to BIG Data.]
Establishments that collect BIG Data don't solely examine what goods their customers buy in their stores or websites; they often connect these purchases to far broader and more detailed information about these customers, so they can improve their marketing, sales and operational efficiencies. These inter-connected, multi-faceted data illustrate the potential power and concerns of unleashed BIG Data covering the countryside – and us.
So far, the big winners in BIG Data are the companies who have the fiscal and staff resources to supply and use it to their corporate advantage. Other beneficiaries include consumers who are resourceful and advantaged enough to use what BIG Data is directly offering customers – discounted prices. So far, this is folks who are card-carrying, digitally connected consumers.
Here are two (2) examples of how BIG Data is allowing retailers to micro-personalize their marketing and sales to selected customers (often, but not always, those who are enrolled in their loyalty card programs). I've labeled these marketing and sales strategies BIG pricing, which is most definitely not like goods' pricing described in economics textbooks.
1. By linking purchasing information to socio-demographic and economic data retailers like Target can make offers to customers that are specifically tailored to their current or expected situations. Large retailers have been using BIG Data processes like predictive analytics for quite some time. It may look simple when you review what coupon offers you receive right after you've bought some items at a retail store like a grocery or clothing store, but predictive analytics is anything but simple. This article, "How companies learn your secrets," describes how Target and other retailers provide purchase suggestions to certain customers in their marketing materials (often coupons), based on their recent purchases and supplementary characteristics, for items a customer wants before they even know they want them.
Supplementary data that can be linked to your purchase history include the location of your home (and when you bought it), age, ethnicity, income level, where you went to college, what kinds of topics you talk about online, brand preferences for certain retail goods, your political leanings, reading habits, charitable giving and the number of cars you own.
Not surprisingly, some customers have taken offense; like the father of a high-school girl who received coupons for baby clothes. He complained to the manager of his local Target store about the coupons, saying, "Are you trying to encourage her to get pregnant?” Later he called back to apologize, telling Target that he had just been told by his daughter that she was pregnant. In other words, Target knew before her father that she was pregnant because of its BIG Data-based predictive analytics. Talk about chilling and creepy. There is nothing random about the sales offers customers receive. Offering lower prices (through customized coupons) for specific items to specific customers based on the seller's BIG Data analyses now happens all the time.
Here's another example of BIG pricing.
2. Safeway and Kroger have inaugurated personalized pricing in some of their stores. Hoping to improve traditionally-thin profit margins, they are creating customer-specific offers and prices, based on shoppers’ behavior. These offers could encourage them to spend more, say buy a bigger box of detergent or bologna if the retailer’s data suggests a shopper has a large family (and more expensive bologna if the data indicate the shopper is not greatly price-conscious). In other words, the price for bottled water that you pay may be lower than the price the shopper right behind you in the check-out line pays – because the retailer knows from your and the other person's shopping habits and history that that unlike you he/she already buys that brand of water (and doesn't need that discount nudge). Personalized pricing is bound to spread.
As I mentioned before, BIG pricing has very little in common with market pricing as described in microeconomics texts. The (singular) "equilibrium price" is determined at the intersection of the market supply and demand curves. (Remember the traditional market supply-demand graph from Econ 101? No? Click here.) This market equilibrium price is not what Safeway, Target, Amazon, Kroger or other BIG Data users are now offering. No, although BIG pricing certainly remains founded on the retailer's supply-side cost attributes, the demand-side aspect of BIG pricing becomes atomized to an individual purchaser's features, not collective qualities of all (or major groupings of) purchasers. As one BIG data analyst said, "It comes down to understanding elasticity at the household level." [Emphasis added.] Demand elasticity is economics jargon for measurement of how sensitive change in the quantity demanded of a good by consumers is to changes in the good's price, the purchasers' income or prices of substitute and complementary goods. Heretofore, it has been rare for economists to consider elasticity at such a fine level as individual households; elasticities were (and are still) calculated for broad markets (like the price elasticity of demand for gasoline) or market segments. There simply wasn't rich-enough data to support estimation at such a disaggregated level as the household. That's no longer true. BIG Data users like Target, Safeway and others have added statisticians and other technical staff to make such calculations so their personalized coupon discounts will be used, not discarded by the intended households.
BIG Data retailers are using smart smartphone apps that allow customers to scan the product barcode in the store and learn the price of that product. What the customer might not know is that in real-time such apps link the customer's potential purchase, socio-economic characteristics and exactly where in the store the customer is, to determine what e-coupon is displayed on the phone – this e-coupon price won't necessarily be the same for every customer. An executive at a company that provides such shopping apps illustrates their sophistication when he states, “If someone is in the baby aisle and they just purchased diapers, we might present to them at that point a [coupon for] baby formula or baby food that might be based on the age of their baby and what food the baby might be ready for.”
Like any rationale consumer, I'm in favor of receiving price discounts to entice me into purchasing another product and/or more of it. Many retailers like Safeway and Amazon have long-standing preferential pricing schemes for certain groups of customers – like their loyalty card customers, or for customers of a certain age (AARP members) or profession (teachers and students). But the new world (brave or not) of personalized BIG pricing and predictive coupons could cast a big chill with respect to issues of transparency and equity.
Carried to its logical end, such tactics might wind up euthanizing the single, listed price for a product. The price of a 6-pack of Bud Light at Safeway wouldn't be $6.49 for anyone/everyone; the price would depend on your individual purchase history and socio-economic characteristics. There would be even less transparency of product prices, making it even more difficult for shoppers to find "the best price" for a given product among stores they shop at. This is, of course, an objective of BIG Pricing – to engender further loyalty of the customer to that retailer.
This growing opaqueness of pricing isn't confined to groceries and books; it includes well-noted changes in fare structures for airlines. With the advent of added special fees for everything from checked baggage and carry-ons to aisle seats, the listed base airfare can end up being as little as 60% of the final total cost, after fees, surcharges and taxes are added in.
Another concern of BIG Data is that it further separates the few, really large national retailers that can afford to take advantage of BIG Data from the multitude of smaller, local/regional retailers who cannot.
Are customers benefited? Probably, at least for groceries and other ground-based retail items. As one customer mentioned, she's now figured out how to game one retailer's BIG pricing system. By alternating her purchases between several different brands of a particular product, she's been offered cheaper prices for the product. I wonder how long it will take this retailer to notice this and change the rules for her personalized coupons.
Answers to questions of fairness and equity that surround personalized BIG pricing are less evident and perhaps chilling. Should those affluent and knowledgeable enough to have certain smartphone apps be the only ones to receive specific price discounts on baby food? Should the price you pay for frozen pizza be determined in some way by when you bought (or lost) your home, your income level and/or the number of cars you own? Some observers are uneasy with this. Professor Joseph Turow from the University of Pennsylvania believes the advent of BIG Pricing by retailers should cause shoppers to be cautious because it may limit their choices and power. He says, “There’s a sense of fairness that’s derailed here.”
Maybe John Donne wasn't correct if you're interested in avoiding BIG data. How could you avoid it? At this point, it would probably mean adopting a much different lifestyle and/or location. I think there are two inter-related options. First, become card-less and cash-full; don't use plastic to pay for anything. Entering the barter economy might be an alternative as well. Second, head for the hills, far away from modern, computer-based commerce. Deal with local, cash-only commercial establishments. This means becoming un-connected to modern communications and commercial systems. Maybe Survivalist Magazine could offer some suggestions. I hear the rural parts of the Cape Verde Islands can be quite nice.
So, welcome to the wide, wide world of BIG Data and BIG Pricing – happy shopping.
[1] I emphasize the bigness of BIG data because of the unimaginably large scale of data sets that proliferate in the Big Data universe. Not so long ago, the upper limit of these data sets could be measured in petabytes (1015 bytes – a million gigabytes). Now the outer reaches of such data are edging towards yottabytes (1024 bytes); that's a yotta data.
