As the “Summer of Snowden” winds down, I’m seeing more articles that draw parallels between online advertising and government surveillance. NPR took it one step further by publishing a story that refers to the practices almost interchangeably, and the New York Times randomly mentions NSA collection in the middle of an article on advertisers using cell phone data.
It’s irresponsible to confuse readers about these two very different types of data capture. The government gathers data regardless of the desires of the parties involved. It can reach into every corner of our lives, from phone calls to chat messages and search queries, and has a track record of bending the rules. Online advertisers need the permission of at least one party, either the consumer or the publisher, in order to capture your data, and in the EU they must notify consumers that such sharing is happening. Stories that conflate government activity with digital marketing will hamper innovation, not only making advertising less relevant—and content creation less profitable—but also scaring people away from participating in data-driven advances in healthcare, education, and transportation (pdf). These benefits hinge on consumer comfort with data-driven endeavors.
To understand how different online advertising is from government surveillance we can start by looking at methods of collection and use. If the US government wants information about you, it can either employ a technology like PRISM or use a Foreign Intelligence Surveillance Act order to compel companies to hand over the data, and forbid them from informing you of the disclosure. In some cases the data are passed to the DEA and IRS who then propagate it further to local law enforcement who can use it as long as they come up with a story about why they really arrested you, in order to obscure the true source of data. The New York Times reports that these data have been used to fight everything from cigarette smuggling to copyright infringement—hardly threats to national security.
In advertising, companies typically gather information with the goal of selling more things with lower marketing costs. The data used include search terms, web pages viewed, things you are shopping for, and more recently, information about what you purchase. There is strong self-regulation that requires opt-in for sensitive information and prohibits the use of any data for decisions on eligibility for insurance coverage. The success of data-driven advertising is measured in the reduction of wasted media spending and often results in showing the consumer more relevant advertising.
It’s a stretch that the government would use information from advertisers for spying. The NSA gathers data in transit that is orders of magnitude more sensitive than what’s used to target ads—the existence of the behavioral advertising industry doesn’t make information gathering any easier for the NSA.
The New York Times editorial board, in an effort to state the case for the regulation of behavioral advertising, came up with a hypothetical scenario of how data collected for advertising could cause harm:
information could also end up in detailed individual profiles that could be obtained by government agencies or purchased by employers or banks to evaluate candidates for jobs or loans.
Yet this hypothetical is ill-founded. The Fair Credit Reporting Act requires high levels of transparency around any data used for financial eligibility decisions. The government doesn’t need advertising data because the NSA gathers data that is orders of magnitude more sensitive than what’s used to target ads. Take the recent report from Time that showing that the NSA captured just under 500,000 contact lists a day. That is to say, the existence of the behavioral advertising industry doesn’t make information gathering any easier for the NSA.
To be fair, there are examples of questionable uses of data by marketers. Last year, the Wall Street Journal documented several cases where cookie data was used to implement variable pricing: customers were offered different prices based on their proximity to a brick-and-mortar competitor. Anecdotally, I’ve heard about airline fares increasing after repeated searches. Though not exactly harmful, what’s perhaps most common is the feeling that being “followed” is creepy and annoying.
But the wholesale demonization of behavioral advertising stands to undermine online business models without confirming a key assumption: Are consumers really bothered by the use of data for advertising once they understand how it’s used? A study we’re running at Enliken has found that less than 10% of the data advertisers use bothers consumers and Acxiom found the same result when they opened up their database to the public a couple of weeks ago. Research from PWC (pdf), DMA, McCann (pdf) supports the idea that consumers want to share information with brands they trust.
Wanting to bridge this disconnect between consumers’ understanding and advertisers’ imperatives has led my colleagues and I at Enliken to build a product we’re calling Transparency as a Service. Enliken makes it simple for any marketer to be more responsible and transparent with the data used to tailor and target advertising. Not only do consumers prefer to do business with companies who adopt transparent data policies, but it also pleases regulators.
Transparency allows for a compromise between advertisers and privacy advocates, one that helps consumers make better choices about who they share their information with while still allowing data to be used in a way that has far-reaching social and economic benefits.