In my last post, I aimed to show that even if advertisements don’t cause us to spend excessively or unwisely, they are still costly in ways that we don’t often talk about. In this post, I’ll explain why targeted ads present additional concerns and are even more worth avoiding than traditional ads.
Part 2: Targeted ads are worse
Targeted advertising is the strategy of using consumer data to inform which ads will have the greatest impact on which potential customers. In my last post I skipped over an important fact, which is that online ads, which constitute most of the ads we see nowadays, are almost all informed by data in some way.
Targeted ads work, and this should come as no surprise. If you can use data to guess with any amount of confidence which viewers will want to buy, say, a lawnmower vs. a mixer, this will lead to more effective ad choices than randomly showing all viewers a lawnmower or a mixer. By looking at demographic data and finding patterns, companies can make inferences about their customers’ temperaments, opinions, worries, and aspirations. Virtually any kind of data can be correlated in some way with shopping habits. Of course, what’s best for the advertisers is not necessarily best for the consumers.
“That machine makes men do unnatural things!”
Privacy
The most popular grievance I hear against targeted ads is that they are a violation of privacy. This resonates with me, but that’s only because I happen to hold privacy as a value: it’s something that I view as good, in and of itself. It’s not that I need to use my privacy in pursuit of some other good; I just like it. So, if someone says, “I don’t care about privacy, I have nothing to hide,” I don’t think I can reason them into valuing privacy as much I do. That’s why I want to go beyond privacy issues in this post.
Still, I’ll make a brief note about privacy because there are some who do value it but don’t think it’s necessarily threatened by targeted ads. For many years, the industry’s answer to the privacy issue was deidentification: “We’ll collect data about you, but we’ll handle it in such a way that we can never know it’s you you, just that it belongs to some individual.”
Lately, the concept of deidentification as a key to privacy-friendly data collection is proving to fall short of its goal, and if my last tech conference is any indication, data collectors are shifting away from it as a go-to defense. The new ethical solution is to get subjects’ data conspicuously, with explicit permission, and modestly—taking only the data that’s needed to answer a specific question defined in advance. That sounds ideal to me, but we are a long way off from everyone in the tech industry doing that, much less other industries. In the meantime, companies are not going to ask permission for what they can get for free.
Edge wants
Traditional advertising is considered annoying when it wastes our time-attention on something we don’t want (Part 1), but what if data collection and analysis gives us ads that only show things we want? Could there be any downside to this? Yes, because “want” and “don’t want” aren’t a simple binary. There are edge wants: things we barely want and things we almost want. Targeted ads will likely home in on both of these, and that’s not necessarily good for us, the consumers:
- Almost-want’s: These are things that, after some consideration, I’ve rationally decided not to buy. However, I am not at my most rational 100% of the time. Sometimes I’m tired, or bored, or just not paying attention to what’s in front of me. These are bad times to be presented with the opportunity to buy that one thing that I’ve just narrowly convinced myself I don’t need.
- Barely-want’s: These items normally just sit on a list somewhere until either they become more important to me (and I buy them) or less important (and I take them off the list). Delaying a decision helps us to consider it more completely and ultimately make a choice that we’re happier with in the long term. Additionally, if you’re into investing you’ll know that it’s always preferable to have more money earlier (spending it later), because this gives you more time to earn investment income. And it’s usually no trouble to let our barely-want’s sit on a list for some weeks or months; most of the time we’ll completely forget about them. Unless, of course, targeted ads are continually reminding us.
A nerdier way to phrase this: When a particular option has a high expected value (EV), it’s easy to make your decision (do it), and when an option has a very negative EV, it’s also an easy decision (don’t do it). But when EV is close to zero, you’re in a kind of “valley of difficult choices” where you have to pay more attention to figure out whether it’s positive-EV at all. Most ordinary ads are to the left of the valley: strongly negative-EV deals that you don’t really consider. But more effective targeted ads will move the needle to the right, on average, forcing you to pay more attention to all ads because now the likelihood of good deals is higher.
Randomized priorities
Suppose I do actually want the thing being advertised, but I wasn’t planning to shop for it right away. For example, I was going to look at it two weeks from now, after my Florida trip, and right now I just want to focus on planning my Florida trip. …And then I see the ad for exactly the product I want, and my attention is too hijacked to ignore it. The decisions I have to make in the near future are ordered by priority, and it takes some amount of mental effort to enforce that priority. Targeted ads actively fight that order by taking some random thing I want and asking me to make that decision right now.
False optimism
To me, the most concerning thing about targeted ads is the common view that despite the issues outlined above, they actually have a benefit for us that makes them worthwhile: Giving up some privacy and perhaps buying a bit more than we normally would are small prices to pay for the privilege of learning about products we’re actually interested in. I think I briefly held this view when I was just starting to work in the field of technology. Now, I think it’s delusion: a way to take this relationship—over which we have no control—and re-frame it as a preferred situation in which our own interests are top-priority. It’s the “I-wanted-this-all-along” denial that saves the ego.
I recently heard this idea from a speaker at a tech conference. He was giving a talk on Google Analytics (a service that analyzes web traffic data) and got to the topic of targeted ads. The speaker told us about about a time he saw an ad out of the blue for an upcoming Sufjan Stevens concert. He was happy because he’s a big fan of Sufjan Stevens; he said, “I no longer care that you targeted me, because this is better for me.”
At first blush this seems like a very real benefit. Without targeted ads, this guy wouldn’t have gone to his concert! The part that he was missing, however, was that if you want to know about a particular product, you can apply for that information specifically. You don’t have to rely on advertisers “getting it right” in their quest to learn who you are and correctly guess what you care about. Thanks to the internet, we live in a world where information is extremely abundant and extremely cheap.
Taking back responsibility
Do you want to know when Sufjan Stevens is playing near you? Add him to an artist discovery service like Songkick or Bandsintown and receive emails only for the artists you’ve specified—not for whatever artist the ticketing company thinks they can sell you on.
Want to know about sales on a particular retail item? Set up a price tracker service like camelcamelcamel. Why wait for Macy’s, Target, or Amazon to learn everything about you? And why endure all of their attempted psychological tricks (and the attention costs mentioned in Part 1) when you can just look at the market yourself?
If we have any idea what we want to buy, we can proactively learn everything we want to know about the product itself, its alternatives, and the state of the market. This is a world where people have the means to find exactly what they want to buy; it’s not a world where the company must find us and tell us what we want. Relying on targeted ads is akin to handing over some measure of personal responsibility (namely the responsibility of finding and obtaining the things we want) to a faceless corporation that doesn’t have our best interests at heart.
To wrap up, I see several real problems with targeted ads, and the apparent benefits don’t stand up to examination. So, what can we do about it?
Blocking ads
I’m no expert, but I can share from personal experience a few easy ways to reduce the influence of ads in your life. Some of these apply to targeted ads, and some apply to all kinds, so you can consider this section my resolution to parts 1 and 2 of this series.
-
- Use a modern web browser and get an ad-blocking extension like Adblock or Adblock Plus (counterintuitively, these are two entirely different services). Then, configure it to actually block all of the ads (of course that’s not the default setting; of course). These work very well, even with YouTube ads.
- Get the Disconnect browser extension (or something similar). Disconnect doesn’t block ads, but it blocks tracker requests, which are bodies of code that get sent to your device alongside website data. They request information from your browser or device and then send that information back to their source. This is one way that companies collect data on your browsing activity.
- Set DuckDuckGo as your browser’s default search engine. Most search engines track your search data and then share it with advertisers (through the questionable loophole of deidentification that I mentioned above). And no, private browsing mode (sometimes known as “incognito mode”) doesn’t prevent this. Private browsing mode prevents your browser (the local application on your device) from recording your activity, but it doesn’t stop Google (the server-side endpoint of your web searches) from keeping track of it. DuckDuckGo is a privacy-focused alternative to Google that doesn’t retain your search data; at worst, it can only show you ads targeted to your current search, uninformed by any past searches.
But… is it possible? Can we really live without Google? It took some getting used to, but I’ve grown very comfortable with DDG. Often the top results are the same as what Google would return, and even when they’re different, it’s not clear that they’re worse. Also, DDG has a unique search feature that makes it super easy to redirect your search to back Google, in case you’re not getting the results you want and think Google will help.
- Always unsubscribe from marketing emails. I’ve made this a habit; even though it’s easier in the short term to just click delete, I always find the tiny “unsubscribe” link and click through it. This works surprisingly well. I’ve only had two cases in which I kept getting emails after unsubscribing from that particular address. When that happened, I set up an email rule to automatically forward the company’s mails back to their customer service address with a small note of explanation; then the marketing emails quickly stopped.
And don’t ask me about automatic spam filters: Gmail mistakenly buried an important career-related email of mine for weeks, and I’ll never trust it again.
- Outsmart Spotify. If you use the free version of Spotify, you can block ads by editing a particular file which filters the data that Spotify can send to your device. This strategy tends to go back and forth in its effectiveness: Spotify will always be looking for new ways to get their ads through, and folks on the other side will always try to adapt their filters. A more consistent solution is to automatically mute Spotify when ads are playing. I use a small program called EZBlocker for Windows; apparently there are similar solutions for Mac.
- Abandon television? I don’t have cable TV, and whenever I happen to see it I’m always shocked at how much airtime is devoted to ads nowadays. Digital streaming services (without ads) provide me with all of my TV-related news and entertainment. But if cable TV is part of your life, then this switch may come at a significant cost, and for that reason I don’t recommend it as strongly as the other tips.
Conclusion
Targeted ads present unique problems for us consumers; they often involve a breach of privacy, they facilitate impulse spending on our “edge wants,” and they threaten to shift the responsibility of choosing purchases from the consumers to the companies. This all comes in addition to the general costs of having a large number of ads in our lives.
The good news is that we don’t need targeted ads to stay informed on the products we care about, and we can take steps to significantly minimize the presence of ads in our lives. In the next and final post of this series, I’ll share some thoughts on what kinds of ads, if any, might be considered good and worth keeping around.
