June 1, 2011
Q&A with Eli Pariser, author of ‘The Filter Bubble’
Eli Pariser‘s new book The Filter Bubble: What the Internet Is Hiding From You is a warning about the perils of Hyper-Personalization (one of our 10 Trends for 2011): As tech giants and websites collect streams of data on users, algorithms that tailor content are sparking “an invisible revolution in how we consume information.” (The Filter Bubble blog includes tips on how to escape this bubble and see the unfiltered Web.) Pariser laid out how this hidden personalization works in a recent TED talk. His book explores the implications, arguing that the online filter bubble is bad for democracy, creativity and innovation. A millennial who has already made his mark in political organizing, Pariser served as executive director of MoveOn.org and is now president of its board.
He talked to us about implications for brands and consumers when choices get edited out of the digital landscape and the counterintuitive opportunity that may lie in giving the people what they don’t necessarily want.
Give us a quick snapshot of how you define the filter bubble and what your take on it is.
It used to be the case that when we went to a website like Google or Yahoo! News, everybody saw the same thing. And increasingly that’s not the case. Your Google results and mine may diverge pretty significantly based on the information that Google has collected about us. And this trend of personalization is sweeping the Web—it’s part of nearly every major website.
It’s concerning for a number of reasons. Partly because it’s happening invisibly, so you don’t see how your information strands are being edited. Partly because there are things that get left out that are important—there are things that you don’t necessarily click a lot but you do need to know. And partly because for a democracy to work, you need people to be able to understand other people’s points of view, and that’s difficult when your own point of view is following you around wherever you go online.
So the filter bubble is the personal, unique universe of information that is generated by all of these filters, all of these algorithms, and that I think we’re increasingly stuck in whether we want it or not.
One of the interesting points in your book is that the bubble will increasingly follow us into the real world. Can you describe how that will happen?
The algorithms that have been at work for a while to target ads are now helping direct where we go and what we do. OKCupid essentially takes the same logic of personalization and applies that to dating; Yelp makes recommendations about where to eat based on this personal information. And so even when you’re going around in the real world, increasingly it’s sort of a function of these algorithms that are trying to figure out what is most relevant to you and pointing you in a particular direction. And the danger here is that to really make good decisions, you need to know what your choices are. And this can make your choices hard to see—it can edit them out of the picture.
The phone is an incredible platform for personalization because it knows exactly where you are and where you’ve been. It has a lot of your Internet data on it. It even knows how fast you’re moving. And so it’s a prime place to incorporate this kind of code, and indeed that’s happening. One of the interesting things about the Apple location leak is that in some ways the location trail that you leave behind says as much about you as any demographic data point. If you know that I woke up in Brooklyn, took the subway to Manhattan, went to Starbucks and then went to work and then went to a movie, you know a lot about me without me ever having to have said any of that stuff explicitly.
And then if you keep being pointed to the same things you’ve always been doing, you’re kind of stuck in that routine, even though you might be really interested in other things—but you don’t know about them from your phone.
Right. So, talking to the folks at Hunch, which does personalized recommendations, they call this “the Chipotle problem,” which is that if you’re doing restaurant recommendations and you’re not careful, you’ll end up always recommending that people go to Chipotle, because people generally have a pretty good experience there, it’s close by most of the time, most people like it. And so you can easily end up just getting redirected and redirected to this one chain restaurant instead of finding out about the more interesting, better, quirkier restaurants around the corner. Chipotle has a very strong signal, and those restaurants might not.
Obviously marketers benefit from targeted marketing, but what’s the downside of Hyper-Personalization for brands?
There are a couple of challenges. One is, as soon as this all really moves into a personalized space, as soon as that’s how people are experiencing a lot of their TV and the rest of their media, which is happening very quickly, the sense of having one big cohesive brand that means the same thing—that gets much harder because you’re wanting to tailor to smaller and smaller groups, and people expect to be tailored to in that way. And so it could be a force for fragmentation of these big brands, and it’s unclear exactly how that all plays out.
Part of the way they pierce the filter bubble—that you get inside someone’s information ecosystem—is you learn these algorithms and you highly optimize, and that’s much easier for big companies than small ones. So, for example, Facebook advertising—initially people thought, “Oh, this is going to be this huge boon to small businesses, they’ll finally be able to reach their target audiences.” That’s been true to an extent, but actually to a much larger extent, it’s allowed Starbucks or other chains to get really, really good at this kind of targeting because they’re working at levels of scale and at economies of scale that small businesses will never reach.
Do you think brands or businesses have a responsibility to open people up to new products or opinions or points of view or whatever the case is?
I think if you’re in the business of organizing and tending to the flow of information, absolutely—you’re helping people get information, then you’re part of the critical information circulatory system of the society, and we need to make sure that the right ideas are circulating around, and that it’s working. In terms of more consumer-facing brands, I don’t know that they have a moral imperative to do it, but I do think that part of what the next step is, there’s a premium on being exciting and unexpected and taking people out of this comfort zone that they’re building for themselves in ways that are comfortable. And I certainly think if you can figure out how to do that, that’s going to be more and more appealing as more and more of what people see is very homogenously tailored.
Can you get into some of the ways in which the filter bubble is changing the consumer experience? For example, in the book you mention websites that could morph to suit the user’s personal tastes.
So we’re used to thinking about this as sort of “If you liked this product, then you’ll like that product. If you liked Lord of the Rings, then you’ll probably like either the Iron Giant or whatever.” But increasingly this is happening not just in what products websites show but how they pitch them and present them. There’s no technological reason why you couldn’t show people a website that had a different design depending on what kind of consumer you thought they were. And increasingly there’s a growing interest in this thing called persuasion profiling, which is essentially figuring out—on an individual basis—are you someone who responds to appeals to authority or to peer pressure or to scarcity, and then building a profile of you that can be used to target you with pitches based on what seems to work. [Pariser recently wrote an essay on persuasion profiling for Wired.]
Are there any other ways in which consumers might be worse off in a personalized bubble?
One of the big ways is that a lot of the most life-changing, transcendent experiences are not the things that will be optimized by most of these algorithms. They’re not the things that you click first. I joke that you need a “It was a hard slog at first, but then it changed my life” button. Because when I think about it, a lot of the things that I love the most, it was really hard going, and I put it down and I came back to it, and then it was amazing. But if you’re living in a world that’s only showing you the stuff you’re most immediately likely to click, you’re not going to see a lot of that stuff, it’s not going to be highly prioritized. And so it would be really sad if that experience stopped happening as much because a few companies were obsessed with getting people to click on stuff once.
I guess the flip side is that people are so used to having everything at their fingertips, they just don’t even have the patience to sit with something and slog through it?
I’m always a little frustrated with when you start talking about “People are this way or that way with their media.” I think the truth is that we all have a more compulsive, more “I want it now” side of us. And we also have a more aspirational, “I want to know about the world, I want to be a good citizen, I want to be well-read” side of us. And different mediums create different balances between those two forces. And what personalization does is it makes it really easy to let the sort of temper tantrum-having, 2-year-old, “I want it now, I want what I want” side win. It helps that side more than it helps the “learning about new ideas” side.
So in your TED video you say we need new gatekeepers; who do you think those gatekeepers will be?
Well, they are right now Google and Facebook and these other companies that a huge amount of the traffic on the Internet runs through. And I don’t know if it’ll be those same companies in the next five years, but I think it will be some companies like that. And just like the media industry, they can choose just to maximize short-term shareholder gains, or they can choose to actually look at the long term and building a society where people are well informed and citizens have the information they need. And I think not only is that better for the country, but it’ll be a more satisfying experience for users of those products in the long run.
Do you think people are starting to lose trust in those companies?
Certainly Google isn’t seen as quite as purely benevolent as it was four years ago. It’s hard for any company to come onto the scene and sustain that for long. But I think the challenge with this is that what they’re doing is hidden, in a way; it’s hidden in plain sight. You don’t know how filtered your search results are, you don’t know in what ways your Google news page has been adjusted. So it’s hard for people to evaluate a service that they’re not aware they’re being provided.
So how do you think brands can open people up to more content or choices or things that challenge them—especially since people are already so overwhelmed by the amount of information that’s out there?
This is what great newspapers do, certainly. And it’s just a matter of building in some other values into these sorting algorithms besides just this very narrow idea of relevance, “that’s what I’m going to click next.” There’s a website called LibraryThing that has something called the UnSuggestor, which is “If you like this, you probably won’t like that.” It’s sort of a silly idea, and indeed the top recommendation is “If you like Critique of Pure Reason by Immanuel Kant, then you probably won’t like Confessions of a Shopaholic,” which seems pretty accurate [laughs]. There are ways of tweaking this that can actually be really fun and interesting.
The tradeoff is, on the one hand, you show them personalized stuff and they go, “Oh, you’re just like me, you have all the same interests I do, I love that.” But the other piece of value that you can give people is the feeling of “Oh, this is a really great place to be informed and to see the big picture and to know what’s going on.” And that experience of inspiration and of serendipity and coming across new things, that’s also a very powerful experience that can be tapped into.
So do you think that long term, there’ll be a countertrend where people really value opportunities to encounter random stuff or different points of view?
I think there will be to some extent. I think as this becomes more pervasive, there will be an increased appreciation for people who do it well versus people who do it poorly. You can have a conversation with some people now where you say, “The iPod Shuffle algorithm is bad,” and people know what you’re talking about—it keeps looping back to the same songs even if you don’t want to hear them and all this other stuff.
So as people get a kind of intuitive sense of how these algorithms are working, they’ll appreciate people who do this well and who do offer some serendipity and who do offer some opposing points of view and some new ideas, and aren’t just reflecting back to the same old thing. I think there will be a lot of people who are looking for that.
Do you think there are certain types of people who are more vulnerable to being isolated in a filter bubble and other types of people who always seek out diverse viewpoints?
That’s probably true to some degree, but again I think actually we all have that in us. In other words, it’s not a matter of there being some people and other people—I think there’s a part of all of us that likes to be pandered to, and there’s a part that likes to see the big picture and get an objective view. And you can focus on one or the other, and you can pull out one or the other.
You talk about inspiration and creativity being tamped down in the filter bubble. So where does inspiration come from in a more personalized world?
The concern is that it doesn’t, because essentially where new ideas come from is a mix of relevant and entirely irrelevant information. And that’s why so many insights come to people in the shower or on a jog or whatever. It’s sort of the combination of an idea that actually an algorithm would never find relevant to the particular problem because it seems like it’s in a totally different sphere. Often that’s what sparks breakthroughs. And so I think the danger is if you’re a physicist and you’re just getting news about physics, you don’t get the news about food that actually inspires you to think about something in a different way.
Photo credit: Jen Campbell
Photo credit: www.thefilterbubble.com