We live in a data-driven age, and much of that data is gathered through our online behavior. Most consumers are aware that accepting cookies on a website or allowing an app to track their location means giving firms more insights into their habits in exchange for some benefit—for example, a helpful recommendation, a personalized discount, or better functionality.
For those who share their data, they face a trade-off: better recommendations but a higher price. But for those consumers who do not share their data, they don’t get the recommendation, and the price is still high.
But new research by Yale SOM’s Jidong Zhou and Andrew Rhodes of the Toulouse School of Economics shows that this choice is not simply a personal trade-off between privacy and personalization. Once a critical mass of consumers shares data, firms begin to respond in ways that could harm those who opt out, leading to an overall decline in consumer welfare—what they call the “privacy choice externality.”
“Every day, we are making a decision about whether we should allow firms to view our browsing or purchase behavior, or whether to use an app that knows a lot about us,” explains Zhou. “Your privacy choice can affect the whole market, not just your own experience. And that effect is often negative.”
Their working paper develops an economic model of privacy and personalization that includes firms’ strategic responses to customer choices—among the first to do so. In Zhou’s view, this two-sided approach offers a more accurate view of the market than assessing consumer behavior alone.
What do those responses look like in practice? Zhou gives the example of looking for a sofa on the home furnishings website Wayfair, a search that brings up over 30,000 items. “If Wayfair has my data, perhaps they can recommend a sofa that will fit my taste better,” sparing him the pain of scrutinizing tens of thousands of potential options, he said. “That should be a good thing, because I can save time in finding a sofa that I like.”
Wayfair, though, is not a direct seller, but rather an intermediary through which many furniture sellers offer their products. If those sellers know that Zhou’s personalized sofa recommendation means he is less likely to browse other options or search further, then, he says, “they will feel less pressure to compete on price,” harming other buyers who have not shared their data.
“For those who share their data, they face a trade-off: better recommendations but a higher price,” Zhou says. “But for those consumers who do not share their data, they don’t get the recommendation, and the price is still high.”
Companies often give us the opportunity to trade data for personalized discounts. If you download the Home Depot app to your phone, Zhou says, “from time to time, you will receive some so-called exclusive coupons that are just for you.” But in order to offer those discounts, Home Depot may raise its “list price,” or public price, in order to knock off a meaningful percentage for its loyal app users. “When more consumers are using the app and sharing their data, then retailers have an incentive to raise their list price,” he said. “Other consumers who do not share data will suffer.”
Some firms even offer customers bespoke products based on the data they receive about their needs or preferences—for example, a healthcare company can create a regimen for specific medical or health conditions, or a clothing company designs a custom garment or curates a selection of clothes. That customization likely comes at a financial cost. To justify the pricier custom product, a company may make its standard products less appealing, says Zhou: “In order to make you more willing to accept those personalized products, which usually have a higher price, the firms have an incentive to make their off-the-shelf products worse”—again, to the detriment of anonymous customers.
The researchers also modeled potential fallout from policies such as the European Union’s General Data Protection Regulation, or GDPR, which can make consumers feel safer sharing data. “If we make data safer, and a data breach is less likely to happen, that’s good for consumers,” Zhou says. “But in our model, that encourages more consumers to share their data, and prices can go up, so other consumers who didn’t share will be harmed, and aggregate consumer welfare can go down in the end.”
In theory, more firms entering a given market helps consumers by encouraging firms to compete on price. But once data-sharing and personalization is introduced into the picture, more competition makes data-sharing more likely—think of how much easier it is to choose among 15 personalized sofa options than to find the perfect one among thousands of options—and that can lead to negative price externalities once more.
That doesn’t mean sharing your data is inherently a harmful act—a Home Depot coupon is nothing to be sniffed at. “We are not claiming that personalization is always bad,” Zhou says. “Personalization often creates good value. But the problem is hidden spillover effects.”
“The Yale School of Management is the graduate business school of Yale University, a private research university in New Haven, Connecticut.”
Please visit the firm link to site