I’ve been thinking recently about advertisers and the internet consumer’s apparent disdain for them. The Pew Research Center recently highlighted consumer distrust of advertising companies, (admittedly, the consumers distrusted just about everything), with the vast majority of consumers stating their distrust stemmed from the use of personal data to target advertisements. Yet despite this apparent consensus among consumers, targeted advertising is alive and well. This disparity between what the consumer “wants” and what they get isn’t unheard of, (see factory farming, Chinese labor rights, the continuing prevalence of Nickelback), but I find the case of targeted advertising unusual because it strikes me as a complete non-issue in a world replete with things to be angry about.
To begin, I suppose I should disclose my biases. I am entirely in favor of targeted advertising. This is the reason why the internet is overwhelmingly free: companies can offer free content and services by tracking user data and providing this information to advertisers. This is the business model of Facebook and Google, and I’m told those sites are doing alright. Even among websites that rely on user generated content for their appeal, advertising pays for the infrastructure that facilitates their interconnectivity. Yet I recognize that many people have an inherent unease with this system, and so I think it’s worth delving into the topic with a bit of vigor to address some of these concerns.
I will assume for the purposes of this discussion that it is not advertising per se that is the problem, and not even targeted advertising specifically, but rather the targeting of advertising based on personal data. After all, in TV, radio, magazines . . . basically every medium where the costs aren’t born by the consumer, advertising is the financial support. And indeed that advertising is highly targeted. There’s a reason that all of the advertisements in men’s magazines are for watches and cars: the magazine has figured out who is buying their product (men) and provides that information to advertisers. The advertisers then select where to purchase advertising space based on their target demographic. Children’s TV shows advertise to children, the Home and Garden channel doesn’t advertise to me, you get the picture. So to some degree, all advertising is targeted.
And indeed the internet works in the same manner, but while before the demographics were projected based on the content, now advertisers can advertise directly to users based upon their internet browsing patterns. So in my case, advertisers look at the content I tend to view (video game, fitness, and pseudo-intellectual articles, mostly) and speculate that I’m a mid-to-late twenties male, so they advertise accordingly (lots of beer ads). And it should be noted that calling this a profile doesn’t necessarily mean that advertisers think or care that I fit the specified demographic. Whether I am actually a mid-to-late twenties male is mostly irrelevant; what matters is whether I’m interested in the endless beer ads they have to offer (not even remotely).
Naturally the distinction between previous forms of targeted advertising and the current trend is that now advertisers advertise to you specifically, rather than a statistical approximation of who you are likely to be. To many, this shift from impersonal ads based on content to personalized ads represents a substantial invasion of privacy. Why?
As always, there is the creepiness factor. I’ve discussed the creepiness factor before, so I won’t go into detail, but there always is the possibility that a negative response to a new technology is an instinctive reaction to things we don’t understand, and not grounded in any substantive privacy concerns. Targeted advertising seems to fit nicely into this framework: it takes something we are all familiar with — advertising — and adds an unusual sense of personalization. (See also the uncanny valley.) And while we haven’t progressed to the point of the Minority Report ads (which address you by name, as if an old friend) the current practice of altering ads based on your viewing habits is, in a word, creepy.
But I’m not sure that the distrust of targeted advertising stems from creepiness alone. After all, it’s been around for a while, and people still seem to dislike it, suggesting something more fundamental. Perhaps the concern is not the advertising per se, but the implications of unchecked data collection generally. Indeed a Pew Survey in 2012 found that 68% of internet users are “not okay” with targeted advertising because it entails them being monitored online. Yet this strikes me as an odd mentality. For one, it’s not like removing targeted advertising would stop online monitoring; some degree of monitoring online is inevitable, and limiting the extent to which it takes place would mostly serve to hamper innovation and worsen the user experience. And for another, data collection also happens in brick and mortar stores. Anyone who has frequented a small business or restaurant probably has created some sort of relationship with the people who work there. (The convenience store near my law school knew my soda purchasing habits by the time of day: caffeine in the morning, caffeine-free in the afternoon). Recognizing patterns in frequent customers isn’t just accepted, it’s expected of a good business. Although we may not like to admit it, we cannot detach ourselves from the world we live in, and the more we interact with the world, the more it is likely to notice.
So maybe the problem isn’t the data collection, but the person doing the collecting. With small businesses, the collector was a person, and people can be held accountable and have natural moral filters. Modern collectors are primarily corporations, where social norms are more difficult to enforce, and where profits tend to trump morals. And while the clerk at your local convenience store can collect some data, corporations can collect everything.
I think this is the primary reason why people dislike targeted advertising: it is a constant reminder of the vast inequality in bargaining power between the consumer and the data collector. Internet consumers know very little about what information is collected about them, and know even less about the company that is collecting that information. Although privacy policies ostensibly provide some transparency, these are often incomprehensible to the layperson and amount to little more than a guarantee. And regardless of whether the consumer knows how their data is used, the end result is one where the company is in a substantially more powerful bargaining position, as they have all of the data. This inequality kindles fear of data abuses, which is facilitated by the lack of meaningful transparency and limited understanding of corporate data usage.
This method of thinking much better aligns with my own: targeted advertising isn’t the problem; it’s a reminder of the problem. Orwell might have described the ubiquitous “Big Brother is watching you,” but in our world it’s “Big Advertising is watching you.” And while Big Brother’s surveillance enforced strict social conformity, Big Advertising’s surveillance is to better sell you stuff. It makes me wonder what else Orwell was almost right about; perhaps 2+2=5 is a metaphor for artificial sales (4? what a bargain!) and doublethink only refers to our bank accounts. I tend to be flippant with 1984 references because I find its overzealous simplicity intellectually dishonest, and while our Big Advertising may present challenges, they are rarely as monolithic and uncomplicated as Big Brother’s.
I’ll confess that while writing this post, I found it surprisingly difficult to stay on advertising, because my mind continually drew me towards data abuses. Advertising was simply so innocuous, it couldn’t really be the problem. I suppose there are arguments to be made against advertising generally, (in the vein of anti-capitalist or anti-commercialist theories), but these have little to do with the privacy concerns people are actually raising. Whereas data abuses provide concrete, redressable privacy harms, and it’s on these that we should be focusing, not advertising. So what are these problems?
Rather than attempt a full list of all potential privacy harms, I’d like to focus on one that I find particularly insidious, even if not terribly prevalent: price steering. Price steering is where businesses alter their prices based on the characteristics of their customer. If a company believes that the customer is wealthy, the company might try to charge that customer more, knowing wealthy customers are willing to pay higher prices. Or if a customer continually searches for flights on a specified date, the airline might raise the price, knowing they are more likely to pay. Price steering builds on the inherent association between online retailers and brick and mortar establishments. In brick and mortar stores, the price is effectively immutable, and even if the price were changed, it would be visibly changed for everyone. We unconsciously transfer these properties to the online environment, assuming that the prices we are seeing are the prices everyone else is seeing. And while competition does serve to combat this to some degree, susceptibility is increasing given the prevalence of mobile purchasing, where browsing multiple retailers is impractical and where the retailer often railroads the consumer into using their purchasing app.
Apart from its overtly discriminatory nature and the visceral unfairness, I chose price steering because it is so clearly difficult for the consumer to effectively police the industry. Unlike discriminatory hiring or identity theft, where the harm is felt immediately, price steering could very easily go unnoticed by all consumers without some form of oversight. And there is no way to solve this problem at the collection stage, because some degree of data will always be available about the consumer, and some degree of collection will always be required to facilitate the transaction (e.g. they cannot deliver a product to you without knowing your address, and your address strongly indicates your socio-economic status, race, etc.).
Despite rampant speculation, I don’t think price steering is particularly widespread. The few studies that have been done suggest at most minor variations in pricing, and industry responses suggest they are price testing, not price steering. This means that if some users saw lower prices it was because the company was testing if the product sold better at that price (and those price tests were always lower). The more common data use was product steering, where companies adjusted the order of search results based on the customer’s profile. (Such as search engines showing slightly more expensive results for Apple products, assuming Apple users were generally wealthier.) Product steering strikes me as substantially less worrisome: we might see the options in a different order, but we all still have the same options.
I think this is indicative of the data collection and uses associated with targeted advertising generally: while it’s easy to speculate about potential abuses, the reality is mostly innocuous. After all, the primary goal of these companies is just to sell you more products. So while harmful practices like price steering should certainly be unlawful, these are the minority of cases, and should not serve as restrictions on data collection or usage generally.
Data collection provides a host of benefits, both to the consumer and the collector, and placing blunt restrictions would be a negative to all parties involved. Data is used to streamline consumer experiences, improve security, and facilitate research, and often the true potential of data isn’t known until it is collected and analyzed. The hype surrounding “Big Data” is not just from what it currently can do, but what it might be able to do. Data is no longer just a part of company business, in many instances, data is their business.
And perhaps more fundamentally, data collection isn’t the problem we should be focusing on. The mere collection of data is almost never intrinsically harmful; what matters is what companies do with the information once it’s collected. The current overemphasis on collection rather ironically breeds a disregard for what is done with data after the collection stage, despite the fact that it is the use of the data that causes the harm.
And while I acknowledge the broader problem of inequality in bargaining power, I don’t think this stems from, or is rectifiable by, data collection. Individuals will never have equal bargaining power with corporations, and it is up to institutions, both governmental and public interest, to advocate on the individual’s behalf. And although I am fully in favor of greater transparency to the individual, the modern world has reached the point where the individual cannot be expected to account for the practices of corporations.
Or maybe people just don’t like ads.