Published on | Prototyping: From UX to Front End — Medium Aaron Weyenberg

Are users equipped to defend against persuasive technology? How much tax on your valuable time do manipulative products take? And where is the border between habit and harm? A conversation with behavioral designer Nir Eyal explores some answers.

“Chains of habit are too light to be felt until they are too heavy to be broken.” — Warren Buffett

I recently observed a couple seated together at the corner table of a popular Colombian restaurant in New York. They weren’t on a date. I knew that because they were not even present for each other. Instead they were each gazing into their phones. Their devices threw a cold glow onto their faces, which didn’t look particularly happy or unhappy. Somewhere in the middle, I’d say.

Later that week I went to a movie at the AMC Village 7 on 3rd Avenue, where two things struck me: The amazing reclining seats, and the fact that AMC Theaters now feels it necessary to include “don’t post” along with the usual “don’t text or talk” part of the theater preroll. It reminded me of when, in one of the more emotionally heavy and somber exhibits of the 9/11 museum, two girls near me snapped selfies. Twice, because one of them didn’t like how she looked in the first one. I don’t know if it was because not being alive in 2001 makes the whole thing abstract history, or because the tribal validation Snapchat provides is too irresistible to delay.

Nir Eyal’s book, Hooked: How to Build Habit Forming Products, tells us some of the causes of these behaviors. In it, Eyal makes a point to separate “habit-forming” from “addicting.” The clearest difference I grokked from Eyal’s writing is this: If the product manipulates us into a compulsion that is said to be good for the our wellbeing, that’s called habit-forming. If it’s bad for our wellbeing, then it’s called an addiction.

Addiction versus habitual use

That distinction between addicting and habit-forming only left me wanting to explore it in further detail. So after Eyal give a brilliant talk about Hooked in New York, I asked him about it during a short Q&A session:

Aaron Weyenberg: Thanks for coming to speak. I thought your book was a fantastic read. Towards the front of your talk you made a distinction between habit-forming and addiction. In fact I think that’s the only time you used the word addiction, in a sort of exclusionary sense. And I made some notes that map to the book. You talk about modulating your mood, habitually using a product with little or no thought, the presence of triggers, reward centers, using [the product] with frequency. This is the language that medical professionals use to… I mean, if you go to an Alcoholics Anonymous meeting, this is the vocabulary that they use. So how does that distinction work in your mind?

Nir Eyal: That’s a terrific question. There’s an entire section in the book called the morality of manipulation.

While Eyal doesn’t avoid the issue of manipulation in the book, that section is less than 1% of the its content. But to be fair, the book is about how to manipulate, not whether you should.

NE (continued): And I knew up front that this was going to be a question. And there’s a very well thought out reason why I didn’t call the book ‘How to Build Addictive Products.’ And I know when someone has or hasn’t read my book when they [say] ‘Oh you wrote that book about how to build addictive products.’ Because I talk about the distinction in the book.

In the first chapter he writes “although some people use the terms interchangeably, habits are not the same things as addictions. The latter word means persistent, compulsive dependencies on a behavior or substance. Addictions, by definition, are self-destructive.”

NE (continued): The difference is, as I said, a behavior done with little or no conscious thought. And we have healthy habits as well as some bad habits. And my hope is that we can use these techniques to build healthy habits in people’s lives. Now, addiction is something else. The definition of addiction is a persistent compulsive dependency on a behavior or substance that harms the user. That key word is “harms the user.” So that’s the difference.

A frequently referenced, broad definition of addiction is a compulsive engagement in rewarding stimuli, despite adverse consequences. Others like Adam Alter add a temporal component. He describes it as “something you enjoy doing in the short term, that undermines your well-being in the long term — but that you do compulsively anyway.” He specifically describes behavioral addiction arising “when a person can’t resist a behavior, which, despite addressing a deep psychological need in the short-term, produces significant harm in the long-term.”

NE (continued): We should not, and can’t for very long quite frankly, build products that harm people. It’s unethical. Now there are some people who get addicted to certain products and we can talk about the psychographic profile and those types of people.

To give an idea of how many people that might be, Eyal writes “Industry estimates for pathological users of even the most habit-forming technologies, such as slot-machine gambling, are just 1 percent.” This figure is cited from the American Gambling Association, whose stated goal is to promote and lobby on behalf of the gaming entertainment industry. The National Council on Problem Gambling shows pathological gambling is over double that rate.

Mark Griffiths of Nottingham Trent University, along with colleagues Nadra Lisha and Steve Sussman of the University of Southern California, found rates of gambling addiction consistent with the NCPG. However, they also found the portion of the adult population suffering from maladaptive signs of at least one addictive disorder over a 12-month period to be an eye-popping 47%.

NE (continued): But for the vast majority of people, you know the problem isn’t that a few companies, like the Facebooks of the world [and] the Instagrams of the world, are building things that people want to use, products that people say suck them in. The real problem is not that a few companies are building products that suck them in. The real problem is that far too much technology out there just plain old sucks.

What Eyal means here is companies that don’t care about your wellbeing aren’t to blame for deploying behavioral manipulation. This non-issue is something Eyal has been saying for years. Virtuous apps that fail to hook you, he argues, is the problem.

Can getting hooked ever serve the user?

Eyal says it can, and gives a snapshot of how habit-forming techniques can help improve our lives.

NE (continued): And that’s why I do what I do, it’s because I know there’s so much potential out there for technology to use these same exact techniques to help people as I said before. To exercise more, save money, be more productive at work, communicate with friends and family. You cannot separate what makes a product engaging from what makes it potentially addicting to some people. And I think the reason I use some of the same language is because it works. And so, I think it works to people’s benefit. Nobody would say that starting a habit around exercising more or eating healthy or being more productive at work is something they wouldn’t, you know, want more [of] or help from technology.

At this point, a nuance emerged; If we accept that software can manipulate behavior, then it seems not quite enough to only ask if that behavior is designed to be beneficial to the user. We should also consider where this behavior is designed to happen, and what role that plays when measuring a product’s success.

Behaviors like those Eyal mentioned — exercising, saving money, being productive with work, communicating with friends and family, a healthier diet—are things that happen away from the product. The techniques researched and taught in places like B.J. Fogg’s Persuasive Technology Lab, and described in Hooked, show us how to keep users maximally engaged with the product itself. The point is to get hooked on the product, not necessarily on the act of saving money or being more productive at work.

“You cannot separate what makes a product engaging from what makes it potentially addicting to some people.”

So do we give more of our time to apps that make us feel happier?

Adam Alter shows us in his 2017 TED Talk that people do report feeling good about using apps for relaxation, health, reading, or education. But he goes on to say people report feeling much less happy about social media, dating, entertainment, news and gaming. That’s maybe not so surprising. But what’s interesting is that we spend three times the amount of time on the apps that we know don’t make us happy. How can that happen? Because those apps hook us.

So I wondered what Eyal thought of those companies that are deploying behavioral manipulation best…

AW: Towards the end [of your talk], you mentioned, and had a slide or two that said [to the effect of] ‘Is this a good idea? Should we be thinking about when to apply these types of techniques?’ And you said, ‘We could use it for good. I’m not teaching Facebook and Instagram how to do this. They already know.’ To my ear it sounded like those companies were not in the category you were talking about.

NE: Not in the category of…?

AW: Not in the category of using these techniques for good. It sounded like you were separating those companies. Did I hear that right?

NE: Well, those are the case studies. They already know these techniques. So I didn’t invent these techniques. I’ve shown people. You know the metaphor I like to use is: I didn’t invent the hammer. I’m showing the physics of how the hammer works and hopefully they can take that hammer and build houses with that hammer. You can also take the hammer and bash someone’s head in, and do bad things with it. And that’s why in the back of the book I talk about the morality of manipulation and how to use this stuff ethically.

I’m a fan of 2 by 2 charts. We see them a lot in pop science and social psychology. What Hooked offers here is one called the Manipulation Matrix, a device intended to help makers find their moral footing. It assigns a question to each axis:

Does the maker use the product?
Does the product materially improve the user’s life?

Nir Eyal’s Manipulation Matrix

In theory, every product should lie in one of the four quadrants. The one labeled “Dealer” is the one to watch out for. It’s where the maker “neither uses the product nor believes it can improve people’s lives. They have the lowest chance of finding long-term success and often find themselves in morally precarious positions.” Still, we are reminded that “with the exception of the addicted 1 percent, users bear ultimate responsibility for their actions.”

Even if users are ensnared by a product made by a dealer, or any other type of maker, Eyal believes users have power over it. Facebook again serves as a case study:

NE (continued): But it’s very hard for me to not sound like I’m defending Facebook. So let me say up front: Fuck Facebook. Like, I have no loyalty to them. I think they provided us the case studies and examples so we can learn from those examples and build products to make the world better. But in terms of people complaining, for example, how Tristan Harris complained… He has an organization called Time Well Spent, that I fully support the mission of Time Well Spent, in that if a product doesn’t serve you, if Facebook is not doing you good, if you feel like you are serving Facebook and Facebook is not serving you, break the habit! That’s why I wrote this book, is number one to help people build habit forming products and number two is to realize, wait a minute, this is being done to me. And so by understanding these forces, you have power over them.

The Atlantic described Tristan Harris as “the closest thing Silicon Valley has to a conscience.” When it comes to who has power over persuasive technology — makers or users —Eyal’s optimism for user empowerment and Harris’ straightforward, if not overtly blunt, pragmatism regarding maker manipulation stand in stark contrast to one another:

NE (continued): The one beef I have with tech critics who have for years been saying ‘Technology is melting our brain and we’re powerless to resist”, and Tristan’s article about hijacking my brain, is that I wish this stuff were that powerful. It ain’t. This is stuff that on the margin helps us do things we want to do. This is not people’s brains on puppet strings. This is not hijacking your brain. And telling people, or you thinking, that you’re powerless, doesn’t serve us. Why? The [inaudible] years ago found that the number one determinant of whether an alcoholic would fall into relapse was not their level of physical dependency on the alcohol. It wasn’t the booze itself. It was their belief about how likely they were to relapse.

Eyal is referring to something called self-efficacy. Several studies do show a relationship between self-efficacy and treatment outcomes. There is one study that found “maximal level of abstinence self-efficacy (i.e., 100% confident) measured at discharge [from treatment] was the strongest predictor of 1-year abstinence.”

NE (continued): And so this is where I agree that we should put technology in its place. I think we should break these bad habits when they don’t serve us. But don’t dare think that we’re powerless. Don’t think that this stuff is irresistible and hijacking your brain. That’s bullshit.

The fact that Adam Alter titled his book Irresistible: The Rise of Addictive Technology and the Business of Keeping Us Hooked indicates another strong dissenting opinion on this very point. The power that Eyal suggests we have over persuasive technology comes from our ability to self-regulate. Self-regulation, though, limits revenue. In Irresistible, Alter quotes Tristan Harris who gives us some bad news: “There are a thousand people on the other side of the screen whose job it is to break down the self-regulation you have.” The playing field is not level. So while Eyal feels power remains with the user, Harris and Alter argue the opposite.

However, Eyal’s call to “put technology in its place” acknowledges at least some degree of imbalance. On this point, he offers ways that it might be corrected:

NE (continued): There are so many things that we can do to regain focus, to put this technology away. There’s things like using new technology, there’s thousands of free apps out there that allow us to modify — you know Facebook is manipulating our brains, there’s no reason why we can’t manipulate them back. So when I use Facebook, I don’t wanna see that news feed. Right? Because I know what they’re using, they’re using these variable reward mechanisms to show me algorithmically driven videos that they know I’m gonna like. So what did I do? I went to the Chrome store, and I downloaded a free Chrome extension called Facebook Newsfeed Eradicator. And it does exactly what it says. When I open up Facebook, I don’t see that news feed anymore. I can go to individual friends, and check in on them. But I don’t wanna see that [inaudible] on my news feed anymore. I don’t use it on my phone, for example. I use it on my desktop then I have time in my calendar to do that.

So I guess that’s a longwinded answer, but it’s all about that harm component. But I ask people, ‘How do these technologies harm us?’ and they say ‘Well it’s distracting me.’ And I say ‘Terrific. Show me on your calendar what it’s distracting you from? Game of Thrones?’ Right? That doesn’t count. Like if there isn’t time on your calendar from something that it’s pulling you away from, that’s the first step. What do you want to be doing with your time? In this day and age when there’s so many distractions, when there’s so many things that can pull us away from what we want to do, if you don’t plan your time, someone else will. And so I think that’s the very first step.

What constitutes harm?

Harm can manifest in many ways. Behavioral and substance addictions alike can harm one’s physical health, mental health, or financial stability. It can interfere with social relationships, hobbies, or work. But there’s something else we all have a finite amount of that’s particularly vulnerable to addictive technologies: Time.

“If you don’t plan your time, someone else will.”

Eyal’s suggestion seems to be that manipulative technology imposes limited harm because it is only siphoning off already poorly managed time, or displacing other equally unfulfilling activities. (It is left to the reader to decide if time spent on an episode of their favorite TV series is better than the same amount of time on Facebook or crushing candy.) But this argument weakens when we see those couples or families, seated at a restaurant table, staring into their devices. Clearly their plans are being interfered with.

Alter articulates this by breaking down a typical workday into components most of us can relate to. We sleep. We work. Then we do things like eat, take care of our families, clean, run errands. (Alter calls these kinds of things “survival”). The leftovers are our personal time. In the span of ten years the share of that time taken up by screens has inverted, leaving us with a much thinner slice. This slice, says Alter, “is where the magic happens. That’s where your humanity lives. And right now, it’s in a very small box.”

Adapted from Why our screens make us less happy. (Adam Alter at TED2017)

Manipulative technology’s constant and increasing annexation of personal time is like high blood pressure. Inconspicuous in the moment, but imminently harmful in aggregate.

NE (continued): And there’s lots and lots of other steps I think to putting technology in its place, starting with removing the triggers, making the action more difficult, delaying the rewards, and making sure we don’t invest. But this is not hard stuff. There’s lots of bad things that these companies do. Sucking us in too much I don’t think is something that we can’t overcome.

Someone else picked up on this thread and asked Eyal about harm in the long term…

Unnamed attendee: When you talked about Facebook, you talked about it’s a place where people go when they’re lonely. There’s also though, from my understanding, people get more depressed after they go on Facebook because they see all these great things going on in people’s lives. So I’m wondering if in the long term it doesn’t have the opposite effect, if it’s going to use it’s efficacy over time in terms of it being addicting, and also if it is actually causing some form of harm?

Nir Eyal: So this is exactly the point. I think if you harm people with a product, you know what they do? They stop using it. Right? After a while they say ‘This makes me feel like rubbish.’ And so that’s going to be Facebook’s biggest problem in the next decade, is if it doesn’t make people feel good, they’ll stop using it. The reward stops being rewarding. And I think we’re already seeing this. The fact that we’re having this conversation.

This is a good time to consider how behavioral addictions work, and how the efforts of the “thousand people on the other side of the screen” create a power imbalance.

Short term reward, long term harm

It’s reasonable to assume a user will stop using a product (or substance for that matter) if it doesn’t make them feel good, but only if that happens in the near term in close proximity to the stimulus. Short term reward and relief from withdrawal sustains the addiction in the face of long term consequences, even when the user is aware of those consequences. In Facebook’s case, by the time you are feeling inadequate about about your own life, you’ve already gotten the reward. The product got what it wanted (the engagement), you’ve gotten what you were seeking (the reward), and on around it goes. This aligns with Alter’s findings that quantifiably show people don’t stop using apps that, in the longer term, make them feel less happy.

In the language of Eyal’s model, Facebook gives you the “reward of the tribe.” In chapter 4 of Hooked, he writes that “a click of the ‘Like’ button provides a variable reward for the content’s creators. ‘Likes’ and comments offer tribal validation for those who shared the content, and provide variable rewards that motivate them to continue posting.”

Do new technologies evolve regulatory systems?

In the case were users still want to use (or can’t resist using) products that have a harmful component, how do they minimize the harm and maximize the benefit? Eyal suggests awareness combined with corrective technologies can bring things back into balance.

NE (continued): We’ve all felt that icky feeling of ‘I’ve wasted all this time with another Netflix episode or spending too much time on Instagram or checking my stupid email when I was supposed to be doing a big project.’ The fact that we are cognizant of this is a good thing. This is the first step to putting this stuff in its place. So I think it’s fantastic. By the way, this always happens after every technological revolution. We had the agricultural revolution give us more food than we know what to do with. It’s the first time in history that more people die of excess food than not enough food. That’s the first time in history that happened. After the industrial revolution, we got cars, we got combustion engines, we got airplanes, and now we have pollution. Now we have global warming. But the solution to each one of these successive technological revolutions, guess what, is more technology! For cars, it’s safety belts. It’s catalytic converters, it’s technologies that improve the environment. Just like when it comes to our personal technology, the cost of this instantaneous connection is that our attention is being fought for and that’s the new commodity — our attention. So we’re going to have to come up with these new technologies and new techniques and adapt to these technologies to put them in their place.

Here are some thought-provoking comparisons to major technological innovations. A pattern is proposed where the innovation creates a problem and successive technologies come in to solve that problem. I think it’s worth examining them in comparison to the innovation of addictive tech.

That people die of too much food is likely a reference to mortality caused by obesity and other diet-related causes. This feels to me not a consequence of the agricultural revolution itself than of the western diet of sugar, sodium and animal products. Culpability belongs with a modern food industry that knowingly manipulates our reward centers. The technological counterweight for this problem remains unrealized, evidenced by ever rising levels of obesity and diet-related disease.

The industrial revolution gave rise to a world economy that runs on fossil fuels, and will continue to for the foreseeable future. Carbon dioxide levels, and global temperatures with them, continue to rise. Innovations like catalytic converters do cut down on emissions of toxins and pollutants, but not carbon dioxide, and so do nothing to curb climate change.

Facebook Newsfeed Eradicator reports 131,000 users. Facebook has more than 2 billion users. For every one feed it eradicates, 15,000 run free. Tools like the Facebook Newsfeed Eradicator are the catalytic converters of behavioral technologies. They don’t have material impact. In all three innovations — industrialized food production, the fossil fuel economy, and behaviorally manipulative technology—the scale of the solutions do not register against the scale of the problems, let alone balance them.

Defending against manipulation

Manipulative software isn’t shooting people, but the idea of technology keeping itself in check is essentially the same one put forward by National Rifle Association CEO Wayne LaPierre, who still maintains that “the only thing that stops a bad guy with a gun is a good guy with a gun.” This approach mitigates gun violence with about the same efficacy as corrective technologies mitigate behaviorally addictive technologies, which is very little.

So how difficult might it be for a third party to get between maker and user? Consider the attempt to regulate Big Soda. Back in 2012, New York City formed an unprecedented public policy and placed it directly between the brain’s pleasure center and a maker (actually sellers in this case) that does not have their user’s wellbeing at heart. A 64 ounce bottle of Coca Cola contains 200 grams of sugar. That’s over a week’s worth of the recommended intake for one adult in a single beverage. Think of it as an analog to the “bottomless bowl” technique of serving content that Tristan Harris describes. So, as part of an anti-obesity campaign, the NYC Board of Health unanimously voted to limit the sale of sugary drinks to 16 ounce containers— still double the recommended amount. It didn’t stick, though. Three years later, New York’s highest court, the New York Court of Appeals, struck down the 16 ounce limit, citing overreach of regulatory authority.

I think Nir Eyal is right when he says technology should be put in its place. The challenge is figuring out who (or what) has a shot at being successful in actually making that happen. Public policy is unlikely to survive an encounter with legal appeals to regulatory overreach. Mitigative technologies in the form of self-regulation apps like Space or the newsfeed eradicator don’t have scale or efficacy. Those with their hands on the levers of manipulation, the makers, will always be heavily incentivized to deliver short term rewards and only get better at it. The only people who aren’t vulnerable to them are those of us without an amygdala or nucleus accumbens, who I suppose have bigger problems.

That leaves users as the only remaining factor, who we know are cognitively underequipped. Eyal himself remarked that “it will be years, perhaps generations, before society develops the mental antibodies to control new habits.” But as the saying goes, whatever remains after eliminating the impossible, no matter how improbable, must be the truth. So what do users have control over?

The role of identity in changing behavior

One thing makers can’t easily manipulate is their user’s identity. Addiction treatment studies show that the social identities people hold about themselves influence relapse. Saying “No, I’m a nonsmoker” versus “No, I’m trying to quit” to an offer of a cigarette can positively influence smoking cessation. So the statements “I don’t use my phone after 9pm” versus “I’m trying to be on my phone less,” or “I’m an ex-Facebook user” versus “I deactivated my account” help shape fundamentally different identities when it comes to relationships with manipulative tech.

Quantifying the manipulation tax

The best defense against manipulation would be deployed where the manipulation actually takes place — in the brain. For that to happen, long term harm would need to somehow be felt way more upstream, near the stimulus. That’s not going to happen outside a lab.

But what we can do is make ourselves aware of the tax, payable in the currency of time, that these technologies impose upon us. If you’ve ever gotten a shocking utility bill, you know how it can incentivize you to conserve. Seamless doesn’t send me monthly rollups of how much money I spend with them. But if they did, I might freak out and temper that habit.

How outrageous a tax you pay is knowable. Apps like Moment can tell you.

The questions users should ask

Just like a tax bill, knowing how much you’re paying is calculable — except it’s adjustable through self reflection in the form of two questions: What do I actually get in exchange for this tax? And how much of it am I willing to pay for it? The couple in the Colombian restaurant may not have asked those questions. If they ever do, my hope is that their faces might be illuminated by the candle instead of two lithium-ion powered tax collectors.

When habit-forming products become addictive was originally published in Prototypr on Medium, where people are continuing the conversation by highlighting and responding to this story.

When habit-forming products become addictive

Cool stuff, every week

More Articles