The data brokerage industry is far more complex and sweeping than meets the eye. Even back in 2012, The Wall Street Journal’s “What They Know” series revealed the extent in staggering terms. In one experiment, a test computer visited the top 50 websites by user traffic. The researchers discovered that their computer was left with 2,224 cookies, installed by more than 131 companies specifically in the business of creating rich consumer-profiling databases.
To say that the public doesn’t know what’s happening is probably incorrect. People are aware of how the commercial infrastructure of the Internet works but either reluctantly brush it off or feel powerless to do anything about it.
Advertisers, employers, bankers, police departments, insurers, credit agencies, schools and hospitals are just some of the many customers of data brokers. And as with any other commercial transaction, these brokers meet the needs of their customers by diversifying the types of products available.
From employers who buy personal data to check on a potential candidate to police departments that profile residents, the sale of data doesn’t just lead to online advertising, it also affects how other social institutions operate. Or take, for example, algorithms used for more effective voter outreach – by analyzing our preferences, deepest wishes, and deepest fears.
Let’s compare our personal data to an agricultural crop. Imagine that on your farm you grow a special crop. All the labor and resources you put into the land returns to you a useful commodity that you can either consume yourself or exchange for something else of value.
Data brokers are digital farmers, except that they just collect the crops of others without owning a farm, and their labor and resources come exclusively from offering free Internet-based services or contracting with companies that do.
We need to view data brokers as rent-seekers: owners of a capital resource that generate gains merely from the sake of owning the resource. In the case of the Internet, this resource is the creativity of the collective commons, that decentralized information-sharing and generating network that now stretches across the planet.
So let’s follow the story of the crop that went to the market and then sailed across the world. Here is the magical trip your data takes.
The seeds you sow
Data begins with those who create it. Online shopping, news, and social media create rich opportunities for data creation.
The labor we input every day – for example, our creativity, knowledge, human interaction, communication – into navigating the digital world are the seeds; the non-stop flurry of images, AI-tailored online content, ad offerings, and TV shows are the water.
Put them together with the soil of the interconnected devices and physical and digital infrastructure of the Internet and data begins to blossom.
Coming to harvest
The data at this point has an exchange value but lacks a direct use-value. But first, it must be harvested.
This happens through a wide range of digital tools that have evolved with the Internet itself. Some of these tools, such as rewards programs, are pretty explicit about what’s happening: you’re exchanging personal data for special offerings and promotional products. Others, like downloadable services that embed data-absorbing software, are less transparent.
Cookies have expanded considerably with the evolution of the Big Data ecosystem. First-party cookies are those that the browser’s address bar shows in its domain and are primarily used for e-commerce purposes like remembering what you have in your shopping cart. Third-party cookies, by contrast, allow tracking from websites and domains from different address bars.
For example, when you download content from a website, third-party cookies allow other websites to “know” what you’ve downloaded. The practice is important for generating patterns in people’s online consumer behaviors.
Third-party cookies form the backbone of the data brokerage industry and allow multiple farmers to simultaneously harvest a massive trot of crops.
Yet the ownership structure of the Big Data economy is less like a privately owned farm and more like a pillaging of the commons.
And then to market
Once the data is collected, it is sold.
Here is where the “ecosystem” of data really takes shape. Hundreds of data brokers find larger data brokers with more resources and aggregated data structures to engage with in commerce. At this point, the data goes in multiple directions. Larger datasets can be sold and aggregated even further – forming a complex and hierarchical chain that ends at some of the largest data-processing companies in the world, such as Oracle, Google and Facebook.
Data can also be processed through functional algorithms and other analytics tools. This “refines” the data and makes it “usable” for predictive purposes. It is only when data is processed by complex analytics, categorization, and predictive modeling that we see the data market reintegrate with the larger complex of activity over the Internet and the digital world generally.
And back home
The massive ecosystem of data-sharing returns full circle after companies that purchase data use it to offer services or products through online advertisements, make adjustments to the platforms or services we use, or predict what we want to see (on Netflix, for example). And this is just one piece of the puzzle. With AI-driven automated decision-making, data takes on a whole new dimension that implicates some of the most important spheres in people’s lives.
However, here the magical trip of our data returns home to us and the process starts anew.
Google recently announced, following other major browsers, that it was going to end the use of third-party cookies. This will not end the data brokerage industry nor the general direction of the Big Data ecosystem.
Rather, by closing off the commons, Google, like other browsers, will have more control over how personal data is generated and collected. This could be both a win and a loss for privacy. A win because other technologies that will phase out third-party cookies could be more privacy-protective. A loss because whatever does replace it may be even less transparent than the cookies themselves.