Personal Data Is Valuable. Give Pricing Power to the People

Winning back trust in the digital economy depends on giving people much more control of the usage, ownership, and value of their information.
Illustration: Casey Chin; Getty Images

Today, people fuel the digital economy with vast streams of data but have virtually no power to demand fair compensation for it. The companies collecting all that data exert full control over the market, raking in billions of dollars without concern for the value of privacy. We need a new market for pricing data, one that can bring balance to this uneven marketplace and establish an economic value for data privacy.

We’ve seen initial discussions on this idea already, with several policymakers and researchers looking for models that would compensate people who share personal data. One such idea, proposed by California Governor Gavin Newsom, would redistribute some of the profits that tech companies reap from user data. Another, from Microsoft’s Jaron Lanier and Glen Weyl, considers the much-needed concept of “data dignity,” which foresees a labor union-like institution that negotiates the terms for data sharing on behalf of end-users. These approaches point toward the much-needed correction of the digital economy’s power imbalances. They provide valuable provocations, but what we ultimately need is a more granular market design that specifies how data value is assessed, secured and traded.

More than anything an effective data marketplace needs a mechanism that accounts for individual users’ privacy sensibilities, which are rooted in personal experience, identity, and specific context. The value of a piece of data changes dynamically based on one’s life situation and how that bit of data is combined with other data points to yield insights. Location data alone might not mean much for someone walking through Manhattan, but adding their shopping, income, and family history could make those GPS details valuable for targeted advertisers. Some data can be cheap and ubiquitous, some are more valuable and worth negotiation. Some might be so valuable that users will withhold them entirely, such as sensitive health data or details about their children. But if we have learned anything from the accumulation of data breaches and disturbing revelations about privacy infringements, it’s that we will not reach the next stage of digital growth without greater trust. That trust must derive from the ability of people to ensure their privacy. Any real economic value of data will require greater user control over data sharing and a higher degree of data scarcity—scarcity that’s created through limits on the non-transparent and uncontrolled proliferation of data transactions. Achieving these aims and, by extension, building a fair and effective market for de-risked data will require four structural elements: personalized data management; the assignment of data ownership; a transaction infrastructure; and dynamic data pricing.

Personalized data management begins to balance the power between buyers and creator-sellers of data by managing personal information and privacy in digital interactions. This requires a fundamental shift in who defines the terms and conditions of such interactions. Today, data management and privacy controls come in the form of either tedious piecemeal offerings or one-size-fits-all solutions. While privacy settings in operating systems, browsers, apps, and other digital services give users some protection and customization, digital service providers still define the scope of those controls. It’s gotten so bad that it now takes about 900 pages and 34 hours to read the terms and conditions of popular apps on an average smartphone. Nine of every 10 users consent to online terms without reading any of them. Other privacy services give users more control, but with blanket solutions that don’t allow them to tailor inherently subjective and personal privacy preferences. As Michael Borrus, a general partner at XSeed Capital, told us, there is no general-purpose solution for individual privacy management. So, for a data market to succeed, individual data creators—whether enterprises or consumers—will need tools to set their own privacy terms and conditions. These tools will need to be able to insert unique identifiers into all kinds of personal data, track those different types of data across various uses and in real time, and negotiate terms regarding usage by third-party companies.

We need to turn the table and at once force and incentivize digital service providers to accept the user’s privacy policy, not the other way around. A “Personalized Privacy Charter” could provide a master control panel for privacy management and begin to establish degrees of scarcity for different types of data in different combinations while also ensuring those data sets against repossession and litigation before they get traded. Such a charter, when combined with the technical and legal tools described below, will begin to create scarcity for any type of data— whether passively or actively created—providing a necessary precondition to negotiate the data’s value.

Of course, no one can sell data if they can’t first establish their data ownership. The most common way to secure ownership of personal data is copyright protection, which provides the owner a bundle of exclusive rights on an original work for the duration of the copyright. However, copyright doesn’t cover certain straightforward facts, such as a person's GPS location, limiting its ability to protect data and create scarcity. While some recent U.S. case law treats data like any other property and provides a basis for protecting someone’s ownership of it, a robust market would need a more effective, modern means.

One brute force approach might be to make data unusable without consent, perhaps by drowning real information in an ocean of decoys. For example, your smartphone could provide your location but then submit a swarm of randomized requests to the same network. A digital service provider almost certainly would deem this a wasteful or even illegal use of its computational capacity, marking users as bad actors and resulting in an outcome no more desirable than the provider’s original sin. A somewhat less adversarial alternative would be a transaction infrastructure that detects unacceptable types of quantities of trackers on a site, and then recommends alternative digital service providers (DSPs) that allow for better protections or agree to negotiate the value of data. This would eventually instigate a privacy competition among DSPs.

Any data market will also need a transaction infrastructure that can recognize market trends, aggregate datasets based on those trends, and then create sellable units of data. Because the value of data depends on so many variables, the very creation of these units would facilitate an orderly framework for the transaction infrastructure to issue sales contracts or data-usage licenses. Such licenses are especially vital, because data is a “non-rival” asset; unlike a cheeseburger or favorite sweater, for example, the same exact data set can be copied multiple times without losing value or quality. Licenses would stipulate the terms for use of the data and clarify the exclusivity of a given dataset. They then could establish a base for the market to enforce those conditions, create an even playing field for market intermediaries, and ultimately foster trust. Of course, this kind of expansive digital economy would require extensive automation of contract and license generation, acceptance, and enforcement. That creates technical challenges, but proponents of data licensing note that it’s possible to solve them.

Finally, the market would need to facilitate dynamic data pricing. Given the existing gap between the actual price of data (about $0.0005 per person, on average) and the perceived value that creators attribute to their loss of privacy (about $36), any shift from the current “free data for free services” model to a data market paradigm will have to happen in stages. This needs to be reflected in how data is priced, because there is no commonly accepted standard pricing model yet. Most current models use high-level economic numbers, such as national productivity levels, and then use that to approximate a fixed value of data per person. These approaches don’t accurately reflect the context in which the information is used nor do they consider factors such as personal privacy preferences—both of which influence the value of different types of data. Hence, we will need dynamic pricing mechanisms that take into account different variables, and do so in as frictionless a way as possible. Since that is difficult when there is no historical data on which to base a more equitable data-pricing system, we could start with a data auction. Buyers could bid on users’ data to establish initial values, which would serve as the foundation for a prediction model that provides more stability by considering a wider range of variables, calculating the probability of matching supply and demand and then establishing prices for different data pools. Again, it’s a complex exercise, but it’s not unlike the complex mix of factors that Airbnb and Uber use in their pricing models.

We have the expertise, algorithms, and computing power to create a new, symmetrical market that establishes a price for privacy-assured data and returns a fair share of that value to the people. The intricate and difficult work needed to build a balanced and effective data market won’t grab headlines. But it is essential to shift the digital economy away from the one-sided objectification and manipulation that has squandered the most important currency of them all—trust—toward an equitable and dignity-based data paradigm that is the foundation for the next phase of digital economic and societal growth.


WIRED Opinion publishes articles by outside contributors representing a wide range of viewpoints. Read more opinions here. Submit an op-ed at opinion@wired.com.


More Great WIRED Stories