As I high school senior, I experienced the power of using statistical models to predict human behavior. At the time, I led a student organization that held events every week or two. Student attendance was recorded at each event. After several months of data collection, I had a dataset on my hands. Analysis revealed how likely students were to attend each event. These projections were based on their RSVP and attendance history on the individual level. Before an event, I would run 1,000 simulations based on each individual’s RSVP status. This allowed me to see a nice bell-curve distribution for the event’s predicted attendance. At the time, I thought that this was what professional marketers were doing with consumer data. That is, using past behavior to predict future behavior. That is partly true.
However, marketers do much, much more with our data. They predict and intervene so they can shape our future behavior. Perhaps most importantly, they know much, much more about our behavior than what we’ve bought and sold. They know which websites we’ve visited. Who we follow on social media. Even where we are in real-time.
The University of Pennsylvania’s Joseph Turow recently examined how brick-and-mortar retailers have increased their data collection efforts in his book The Aisles Have Eyes. These data collection efforts, Turow writes, are a direct response to the success of Amazon. He documents how our phones often tell retailers exactly where we are within their stores. Through interviews with industry experts, he examines the massive efforts that are underway to merge the data they have about their customers with third-party data.
While Turow’s account may be creepy, he does point out that consumers do often benefit from being tracked. When you opt-in to give your grocery store your contact information and let them track your purchases via your loyalty card, you often receive additional discounts personalized for you. Similarly, when you let a retailer’s app track your location, you usually receive an offer in exchange—like 5% off your next purchase.
What consumers often fail to realize is just how much data retailers actually have on us. We fail to realize that even if we don’t have a retailer’s app, they can still find out our location history by buying it from another company whose app we do have.
Further, data from a survey Turow conducted indicates that most people disagree philosophically with the idea that we can trade our privacy for lower prices and personalization.
Consumer Data and Price Discrimination
Perhaps the most important idea that Turow covers in his book is price discrimination. Price discrimination the often well-intentioned economic practice of charging different people different prices for the same product. A positive example of this is higher education, where students from lower-income families often receive a discount on their education in the form of financial aid.
However, as Turow points out, price discrimination can take a form that’s socially unjust. If companies determine that a customer’s income or shopping patterns indicate that they have a low lifetime customer value, this customer will receive fewer discounts. Even more, customers who decided to protect as much of their own data as possible will lose out on many of the discounts available.
Consumer Data is Changing More Than Just Prices
Once companies know what “type” of consumer we are, they may push us to be more like that “type.” For example, let’s say a digital advertising company notices that a given consumer—we’ll call her Julie—enjoys Starbucks Coffee and occasionally goes to SoulCycle. Let’s assume that a digital advertising firm has a pre-segmented “type” of consumer. This consumer frequents Starbucks and SoulCycle, but also subscribes to Netflix. The problem is that Julie never watches television. She occasionally visits the movies with friends. However, when she’s alone she spends most of her time reading.
The algorithms don’t know she reads. Julie doesn’t shop at bookstores, either. She checks out books from the public library down the street from her office. The public library doesn’t sell her data to third-parties. They keep the information about what books she checks out and the very occasional late fee private. In many ways, Julie is a bit of an outlier. Her love of libraries over Netflix helps define her as an individual. Unfortunately, advertisers believe that because Julie purchases her coffee from Starbucks and attends spin classes at SoulCycle, she’s also very likely to subscribe to Netflix. This is not true.
The danger here is that with the right pricing—perhaps a 3-month free trial—and the right messaging, there’s a chance that Julie subscribes. Sure, she could unsubscribe. But there’s also the very real possibility that she replaces her Tuesday night habit of popping by the public library with scrolling through the latest releases on Netflix.
Consumer Data and Preference Conformity
With big data, there may exist a convergence towards social conformity within certain market segments. This is in stark contrast to the promise that the use of consumer data would lead to increased personalization. That’s true, but that “personalized” experience may push us to be more like those within our market segment, stripping us of our individual preferences.
What’s even more dangerous, as Rushkoff notes, is that social networks may have an incentive to intentionally push us to like some of our interest more than others. Returning to Julie, let’s say that she has indicated through her activity on Facebook that she enjoys both music and comedy equally. On an average Friday night, she would have no preference between attending a concert and a comedy show.
The problem is a matter of economic incentives for Facebook. Let’s say that Facebook projects that they can earn $3 in advertising revenue per year from music fans. And let’s assume the company projects they can earn $7 per year from comedy fans. Here, Facebook has an incentive not only to show more comedy-related ads to Julie, but also push more “organic” comedy content on her.
A Search for Solutions
It is clear that there are both benefits and costs to the increased use of consumer data. On the one hand, consumers get a more personalized shopping experience. On the other hand, consumers with lower incomes will likely have to pay higher prices at the register. Additionally, those who wish to keep their data private will also pay higher prices.
For marketers, there is little choice in whether or not to leverage big data. By abstaining from large-scale consumer data operations, they put themselves at a disadvantage relative to the competition. Perhaps the best solution is for marketers to work together to create some kind of certification similar to what is used in the food industry for GMO-free products. Brands could display a seal saying they abide by certain standards to use my consumer data responsibly. I’d trust those brands.