These days, if you’re in business, you’re probably going along with Big Data. You’ve probably been told of its importance by the high priests of tech. You’ve been both scared and awed by its reach. Scared of missing out and a little worried about going down the path and becoming a true believer.
Now might be a good time to elaborate more on this idea of lean data or responsible tech. Call it our inadvertent business strategy. It’s never going to be as big as Big Data – hence its leanness. Let’s get that out of the way at the beginning. We’re not stuffing the Big Data genie back in the lamp.
But Big Data should not be confused with all data.
For as long as WeTransfer has existed, the large companies around us have been talking about Big Data. Big Data was going to be the saving grace of mankind, as we would be able to deliver exactly what people wanted exactly when they wanted it.
It was rightly celebrated. It was, in that unfortunate phrase, a ‘game changer’. But the phrase ‘Big Data’ was often yelled out in a way that seemed creepily evangelical. Tech people have, I’ve noticed, a tendency to sound like unhinged preachers when it comes to the latest find.
The fact that it was an uphill struggle was evidenced by the lack of interest from the majority of major advertisers and media companies because of our lack of data. Because of our inability to target, re-target and profile. Unlike many, we were unable to predict your menstrual cycle. It would take a maverick within an organization, who possessed that rare ability many in media don’t have, to be able to listen to and take a punt on WeTransfer and allow us to make him a local hero, by demonstrating the power of design over data.
The sheer arrogance of the IAB (institute of advertising standards), who refuse to recognize us as a standard format, to this very day, is a demonstration of how we are still being seen as an outsider or challenger, despite our scale.
The contempt we were shown by many of the big media conglomerates when presenting our wares, as ‘they simply couldn’t recommend us’ – we were often told: ‘We’ll try, perhaps X brand has some experimental budget we can find?’ – led us to think we’ll do this on our own. So we did.
And then there were the investors, who would look down their noses at us, as they weren’t able to model without that data. I’m sure no one could model the Model T when Henry was busy revolutionizing the world.
In the mighty emerging world of Big Data, the act of searching for a nearby waffle restaurant could very well be a step towards understanding the secrets of mankind, bit by bit, imprint by imprint.
‘The everyday act of typing a word or phrase into a compact, rectangular white box leaves a small trace of truth that, when multiplied by millions, eventually reveals profound realities,’ Seth Stephens-Davidowitz announces in his book Everybody Lies: Big Data, New Data, and What the Internet Can Tell Us About Who We Really Are.
‘Let me be blunt,’ he goes on. ‘I am now convinced that Google searches are the most important data set ever collected on the human psyche.’
For Stephens-Davidowitz and others, the value of Big Data is its ability to collect data that would otherwise never see the light of day. It would normally lie around untapped, never be admitted or spoken aloud to the world at large.
‘At the risk of sounding grandiose,’ he says, ‘I have come to believe that the new data increasingly available in our digital age will radically expand our understanding of humankind.’
Its proponents are not only enamoured of the amount of data but also what can possibly be considered data. For people like Stephens-Davidowitz, there’s no limit. Take sex. There’s so much data on porn now, the stuff is of the quality ‘Schopenhauer, Nietzsche, Freud, and Foucault would have drooled over’.
Most importantly, this data is, in his word, ‘honest’. Others have used different ways of describing it. This is ‘dark data’, an untapped landscape of the soon-to-be-quantifiable. Believers hint at limitless reach and no reason to ever stop collecting and collecting.
‘In the pre-digital age,’ Stephens-Davidowitz writes, ‘people hid their embarrassing thoughts from other people. In the digital age, they still hide them from other people, but not from the internet …’
It’s the ultimate truth-teller. The implication here is that we should delve deeper and deeper and deeper, but not to worry: there will be a separation. Big Data will harvest all this honest data – and by ‘honest’ we mean our most intimate revelations. But ‘other people’ won’t find out. ‘Big Data allows us to finally see what people really want and really do, not what they say they want and say they do.’
The excitement is always palpable when reading the writings of the evangelical believers. It’s often difficult to find any downside, any threat, any hint of a shadow. In the scenario above there remains some sort of well-policed divide between data and identity. This barrier, it’s implied, will be there to keep us safe. But who is manning that barrier? Is that the barrier whose security and strength is being protected by Facebook, the notorious bastion of unhackable safety? By Google?
As the evangelical word of Big Data spread, I started to notice the way its proponents were excited by their own insatiability. Everything could become data. Everything was already data. Light data, dark data.
‘These days, a data scientist must not limit herself to a narrow or traditional view of data,’ writes Stephens-Davidowitz, ‘These days, photographs of supermarket lines are valuable data. The fullness of supermarket bins is data. The ripeness of apples is data. Photos from outer space are data. The curvature of lips is data. Everything is data!’
The next line, which ends the chapter, caught my attention. ‘And with all this new data,’ he writes, ‘we can finally see through people’s lies.’
I’m not immune to the excitement of Big Data.
We use data to understand who our customers are and what they’re looking for.
We use data to understand where best to locate servers for speed. To ensure that we can guarantee uptime. We use data to understand which type of imagery performs best (as in how many more people click on an image of Madame Gandhi versus Gandhi).
And increasingly we use it to understand how we can offer more creative solutions to our customers.
There is a Japanese philosophy (of course) called Omotenashi. It focuses on human-to-human interaction. The basic principles are as follows.
It’s not that we don’t have access to data, it’s that we choose to use it responsibly. It’s the same as a trustworthy bank. Sure, they have access to all of your money, but do they choose to leverage it all?
The love of Big Data was starting to sound like something else, something larger and larger, and gleefully out of control. I’m not the only one to notice the omnivorous appetite of Big Data extraction systems – in love with their own ability to consume, able to eat up whatever they want, indiscriminate, fat and getting fatter, never satisfied.
As it gets bigger and bigger, Big Data will reportedly tell us more truths. For some, the data reveals us in new ways. For others, such as John Cheney-Lippold, a professor of American Culture and Digital Studies at the University of Michigan, Big Data doesn’t speak our hidden truths. Instead, it actually speaks for us.
Big Data won’t eradicate lying. It’s just that the definition of ‘lying’ changes. We’re flooded with ‘truths’. Algorithms, Cheney-Lippold explains in his book We Are Data, can do anything. They can effectively ventriloquize our data. If you wanted to, algorithms could ‘make our data speak as if we were a man, woman, Santa Claus, citizen, Asian, and/or wealthy’.
These would all be true. This data builds ‘a proprietary vocabulary that speaks for us – to marketers, political campaigns, government dragnets, and others – whether we know about it, like it, or not. Who we are in the face of algorithmic interpretation is who we are calculated to be.’
He continues: ‘When our embodied individualities get ignored, we increasingly lose control not just over life but over how life itself is defined.’
Big Data evangelicals speak of honesty, truthfulness, unguarded data that reflects the world as it truly is. Cheney-Lippold brings more nuance. Our online selves are not just endless oil derricks of truth. Which self do they want? We’ve got a few they can choose from.
For him, this online self ‘to borrow the overused phraseology of pop psychology, is a schizophrenic phenomenon. We are likely made a thousand times over in the course of just one day.’
Disclosure to one rectangular box might offer some interesting stats but the voraciousness of the Big Data project brings about a more complicated reality.
‘Who we are,’ Cheney-Lippold writes, ‘is composed of an almost innumerable collection of interpretive layers, of hundreds of different companies and agencies identifying us in thousands of competing ways. At this very moment Google may algorithmically think I’m male, whereas digital advertising company Quantcast could say I’m female, and web-analytic firm Alexa might be unsure. Who is right? Well, nobody really.’
What’s being constructed is a secondary text that we don’t ever get a chance to read. ‘We have reached the point,’ writes Shoshana Zuboff in her opus The Age of Surveillance Capitalism, ‘at which there is little that is omitted from the continuous accretion of this new electronic text … [which] spreads silently but relentlessly, like a colossal oil slick engulfing everything in its path.’
Who gets to make sense of Big Data? ‘The essential questions,’ Zuboff writes, ‘confront us at every turn: Who knows? Who decides? Who decides who decides?’
Before I get to why it might be good business, let’s just look at what incessant data mining does to the infrastructure of our online world. How might the cracks form? You’re never going to get a venture capitalist to listen to the following (not many), not yet, but hey, the thought is out there. Can we trust it? Can we as business people ask some big questions? Will data mining stop us from developing as businesses? As people?
(Sorry to bring the notion of a catastrophic hindrance of the development of self to your airport reading but, hey, here it is.)
‘When identity is formed without our conscious interaction with others, we are never free to develop – nor do we know how to develop,’ Cheney-Lippold writes. Google might look at your search terms and assume a gender. Is that a big deal? ‘What an algorithmic gender signifies is something largely illegible to us, although it remains increasingly efficacious for those who are using our data to market, surveil, or control us.’
Political economist Oscar Gandy points out: ‘The use of predictive models based on historical data is inherently conservative.’
With this chorus of criticism, will there be a time when Big Data starts to look like a public health concern? Who knows? Should our kids be under surveillance, defined and stuck in proprietary categories from birth onwards?
Or would you prefer a childhood that is somehow more organic?
Here’s Cheney-Lippold again: ‘In an online world of endlessly overhauled algorithmic knowledge, Google’s misrepresentation of my gender and age isn’t an error. It’s a reconfiguration, a freshly minted algorithmic truth that cares little about being authentic but cares a lot about being an effective metric for classification.’
Let’s rein it back a bit. This is, after all, supposed to be a book about making a billion dollars by not pestering your users online. Some say Big Data will reveal who we are (even if we’re lying to our IRL friends). Some say it will tell us, forcefully, unendingly, exactly who we are.
What lurks in the background is another question.
Will people someday get sick enough of both inadvertently exposing their truths and secretly having roles and traits assigned to them?
The ‘right to be forgotten’ sounds a little like a Milan Kundera novel – before it became policy. Maybe the sequel will be the ‘right to control how life is defined’?
Increasingly, people turn to us when they intuit there could be a business opportunity here. Is there money to be made? In a way, the problems we’re seeing with Big Data and its seemingly inevitable importance in our lives reveal a vulnerability. We all may draw the line somewhere.
I’m reminded of a Deloitte study quoted in Zuboff’s book. The subject was auto insurance and the increased monitoring of drivers using telemetry – the process of recording and transmitting information. Deloitte acknowledged that ‘according to its own survey data, most consumers reject telematics on the basis of privacy concerns and mistrust companies that want to monitor their behavior. This reluctance can be overcome, the consultants advise, by offering cost savings “significant enough” that people are willing “to make the [privacy] trade-off,” in spite of “lingering concerns.”’
What if those concerns continue to linger for those confronted with Big Data in all its forms? That’s a big what if.
But the tech world is not immune from people finding out how they’re being used.
Although we valued the data we collected from our users, at WeTransfer we started to realize we were now on the other side of what seems to be the most important divide of our time. Perhaps this process started as just an incomplete and naive understanding of the future of the internet. Perhaps it was because we viewed the world in too simple a manner. Perhaps it was luck.
Back in 2009, we set up a service that trimmed the amount of data we asked for. We had no sign-up. We didn’t ask people to create an account. This might not seem like a big deal.
But this is the single move that became one of the most compelling and groundbreaking aspects of WeTransfer and still sets us apart from the data extractors. It was, some would unfortunately say, our game changer. (I’m never going to use that phrase again.) For better or worse, we were distinguished from the rest. This became the USP that sets us apart from the competition. It became the calling card that separates WeTransfer from nearly every other tech business out there.
If the little rectangular box of a Google search defines its ambitions to collect and compile, our sign-up box hinted at our own burgeoning beliefs:
Leave people alone.
This meant that the amount of data we would collect on and about our users would be very limited. Compound this with an advertising format that couldn’t be sold by the programmatic websites of the world, and we were left with nearly zero data.
Between the years of 2009 and 2015, this was seen by advertisers, and sales houses and brands, investors and corporations as a serious shortcoming. Like: What the hell are you doing? Like: Why not reach out and grab what’s in front of you? Like: Are you guys high on Dutch weed? Look at all the cash you’re leaving on the table.
It was, at best, a missed opportunity. At worst, we were dismissed as tech guys who unfortunately just didn’t understand tech. Which would seem to be a crucial thing to understand.
There was, of course, some degree of truth in this. I can admit that now. None of us were hugely tech literate. None of us came from an engineering school or university where we had learned to code C++, Python, Ruby or any other language. But we had stuff going for us. We could speak multiple languages. We understood the language of design and the experience of design. We firmly understood that our attitude towards sign-up could be a disruption. Maybe not the kind of disruption people were used to. ‘Disruption-by-trusting-our-users’ wasn’t exactly a buzzword or a TED Talk. We could, we realized, disrupt an environment where people were asked to hand over information before they’d received anything in return. This disruption stemmed from a truly crazy, mind-bending concept: actual loyalty.
Perhaps it was our archaic, European view of the world. Perhaps this was our Dutchness shining through again. We couldn’t shake the fact we didn’t need to ditch our own values when we went online. We couldn’t recall ever visiting a physical store in the real world that would ask us to sign up for a loyalty card prior to having purchased anything. We were going to apply the same utterly loony bit of disruptive thinking to the online world. If a store in the real world didn’t ask you, why would any internet business ask you?
As the behemoth that is data extraction charges forward, this choice feels like more of a disruption, year after year. Back then, we were helped by our size. We were a tiny team with no money. The more complex we made the product, the more people we needed, and the greater the cost.
In 2010 we had just enough money to pay a small legal team to advise us. I spent a big chunk of my time working behind the scenes, ensuring our terms and conditions and privacy policy were simple. Lean – as we liked to call it. I spent a lot of time ensuring that, unlike any other business of our ilk, we had dead simple legal speak. Our terms were easy to understand, even our employment contracts. Everything should be set up in a way that was inclusive and obvious. It wasn’t difficult to set ourselves apart from the rest of the file sharing crowd.
File sharing in 2009–2010 was dominated by businesses that were mostly interested in piracy and porn. We wanted to ensure that we weren’t. We held our head high and stayed clear of the pirates. I joined the Notice and Take Down Policy Group in 2010, to ensure we were actively involved in bodies that were out to protect copyright, prevent piracy and deter misuse on the platform.
All the while our lean data policy was supported by one of the best customer support teams in the world, constantly helping to keep users happy, being on call 24/7 and escalating issues when necessary. Shout-out to customer support! They’re game changers! (I know, I said I’d never use that phrase again.)
There was more evidence of our wild appetite to disrupt: you didn’t need to share the details of your life with us to get customer service. As we all know, the internet can be a dark place. We started to notice that where there were huge amounts of security, encryption and firewalls, the internet’s underworld found refuge. When there was transparency and trust, the underworld tended to flee.
Knock on wood.
In the meantime, our simple service has transferred over 100 billion files to every corner of the globe. We’ve noticed the majority of our users are professional and grateful that we provide a mostly faultless, simple service.
Sometimes – like anyone involved in tech – I think about what we could possibly learn from them, what details we could take from them, how interesting it might be to monitor and dissect and scrutinize their moves, their content, their choices. Of course that urge is there. I can understand Stephens-Davidowitz’s enamoured view of Big Data as ‘digital truth serum’.
But Big Data is not a practice or a disruptive tool. It’s an environment. It’s being used to aid a new, rampaging form of capitalism – voracious and looking to suction up endless amounts of our data. That’s not hyperbole at this point.
Some companies will go in with everything they’ve got. We’re already seeing toy companies and insurance companies enraptured by the idea of capturing data. It begins to resemble full-bore mining. Some, without expertise, will get mired down in the world of Big Data, awash in numbers with no real way of discerning what they mean.
For various different reasons, more and more will go lean. There will be a movement towards responsible tech.
For inspiration, we’ve always looked towards some of the aforementioned real-world companies. Being responsible doesn’t mean getting abused or going belly-up. Look at REI Co-op, the place to go for outdoor clothing and gear. How is it that a business as successful and loved as REI can offer an unlimited, no questions asked, full refund, full return policy and not get abused?
The choice is whether you want to live in a world where you expect the worst from people, or one where you expect the best?
I don’t want to make it sound like it’s easy to find allies. Our business development team spends a large chunk of its time seeking out potential partners. They look for companies who might embark on collaborations with us, companies that might share our values and beliefs. During one session, we drew a chart listing all the tech companies that we felt aligned with us.
Two minutes later we took a break.
In the for-profit world the list consisted of: Lyft, Firefox, Omidyar Network, DuckDuckGo, Craigslist and Aspiration bank. There was an incredibly long list of companies that didn’t fit with our values. I’m sure you can guess a few names.
In the not-for-profit world the task was a lot easier. If we considered retail we would include the following companies: Patagonia, REI (as mentioned already), Ben & Jerry’s, TOMS Shoes, SpiderOak.
What was evident was that there was not a single advertising business on the internet doing what we do. There was not a single SaaS business on the internet doing what we do. There were very few venture-backed businesses doing what we do.
Still, it feels like the right direction.
(Fingers crossed.)
Let’s look more closely at one success story. The search engine DuckDuckGo has had a very clear mission from the beginning: ‘Too many people believe that you simply can’t expect privacy on the Internet. We disagree and have made it our mission to set a new standard of trust online.’
DDG pops up all the time on those ‘How to Protect Your Digital Life in 9 Easy Steps’ lists. It always gets mentioned when there’s yet another Facebook break-in. It seems that in the last few years the message is getting through to people who have been burned one too many times. DDG are not interested in shrouding their beliefs in highfalutin ideals or impenetrable tech-speak.
‘At DuckDuckGo,’ their website states, ‘we don’t think the Internet should feel so creepy and getting the privacy you deserve online should be as simple as closing the blinds.’
Privacy should be the default. The company knows, as do we, that users need to be led back. They need to be reminded of what has been so swiftly taken away.
We’re living through a time where people could easily forget the importance of privacy. And instead of castigating the world, like some angry screeching preacher on a mountaintop, there are ways to lead people back.
At its core DDG is a search engine, but they also understand the need to take on the identity of a teacher – a pleasant teacher – with blog posts that seem to be written by their mascot, a talking duck named Dax.
Just recently, DuckDuckGo released on their blog an article schooling their users. It was called ‘Three Reasons Why the “Nothing to Hide” Argument Is Flawed’ and tucked away in the answers were sentiments that would have seemed obvious a few years ago.
‘Simply put,’ read the last line in one reason, ‘everyone wants to keep certain things private and you can easily illustrate that by asking people to let you make all their emails, texts, searches, financial information, medical information, etc. public. Very few people will say yes.’
All this benevolent teaching is fine and good, but DDG have, since the beginning, also held the belief that this approach can work, can make money, and will only become more popular as we’re dredged for more and more data. They’ve been around since 2008 and have, from the outset, been interested in demonstrating there could be money made, eventually, by those who don’t trample users’ privacy needs.
Back in January 2018 they expanded the mobile app and the necessary browser extension. According to their website, they’re up to 21 billion searches.
The answer, for DDG, comes from advertising. Again, like us, they employ a throwback model. ‘If you type in “car” you get a car ad, if you type in “mortgage” you get a mortgage ad,’ its founder Gabriel Weinberg told Wired. ‘We don’t need to know about you or your search history to deliver a lucrative ad.’
There something confident about the ‘we don’t need to know about you’ in the statement above. Perhaps this is the beginning of companies simplifying the relationship, reconsidering what is necessary. For DDG, the transaction is just fine without access to a search history. There’s trust involved.
I mean, you’re still going to see an ad. There’s no way of getting around that.
It benefits DDG to remind people about the severity of the recent scandals. It benefits us to get people to feel this sort of thing as they might feel pain on their skin. There are many in the tech world who would like to underplay the importance of what’s erupted in the last year. The Cambridge Analytica scandal was, according to Facebook, pretty bad. ‘People’s privacy and security is incredibly important, and we’re sorry this happened,’ Facebook said when the scandal erupted. Because they are the only game in town, Facebook is able to perform, again and again, a meaningless ritual of contrition. ‘Because Facebook so dominates social networking,’ co-founder Chris Hughes wrote, in May 2019, ‘it faces no market-based accountability. This means that every time Facebook messes up, we repeat an exhausting pattern: first outrage, then disappointment and, finally, resignation.’
In September, 2018, they announced at least 50 million and potentially up to 90 million Facebook users had their data exposed to hackers. Fifty million? That’s like all of Kenya getting hacked. It benefits companies like DDG and our own to remind users – even employing the skills of an animated talking duck – that this stuff is seriously worthy of more than another ‘sorry’.
The Big Data titans would like us to follow them into the reality of the new normal, where it’s normal to give up everything, no matter how many times we get abused. The new normal is acquiescence. It’s difficult to get some people to forget the old normal: that the responsibility of holding data is serious.
The issues that have arisen through the General Data Protection Regulation (GDPR) go to the core of how we want to live. The aftermath of Cambridge Analytica has rippled through Silicon Valley and, hopefully, into the lives of those who use their products. I’m confident to say this is just the beginning.
We’ll see many more companies follow down the path of responsible data. We’ll pay these companies for their services. We’ll make some people very rich. (The data extractors will inevitably get richer, but still …)
Instead of summing up their words, I decided to speak to a few of our allies on the web.
Gabriel Weinberg is CEO of DuckDuckGo, and after learning many interesting things about privacy from their talking duck, I decided I should speak to a real person. Why not start at the top? Weinberg started his bare-bones search engine at home in Pennsylvania. He was complimented by all sorts of sources. The site was compared by Time magazine to hamburger restaurant In-N-Out. ‘Just as In-N-Out doesn’t have lattes or Asian salads or sundaes or scrambled eggs, DDG doesn’t try to do news or blogs or books or images.’
Weinberg was able to grow slowly away from the In-N-Out comparisons. DDG now offers news and blogs and images in the search engine and considers itself an internet privacy company, not just a search engine, with tracker blocking and smarter encryption along with private search.
Gabriel: I had another company which I sold. Then I was looking for something to do more long term. But really I wasn’t sure what that was, and so I tried to ideate.
Then I realized I was not going to be doing a good job with that.
So, like WeTransfer, I started a bunch of side projects, and tried to see what stuck in terms of my interests. A lot of those ended up being search related: trying to get SEO, building something to remove bad clicks.
The other piece was instant answers, similar to Wikipedia. There were sets of links out there. For instance, Wikipedia external links were often better than Google’s top results for certain things.
So I decided all of those things could then be put together maybe, to make an interesting search experience.
I started playing around with that as a product. My job from 2007 to 2010 was just a lot of iteration. I had a first child at the time. I was a stay-at-home dad, so I was just doing it on the side, and responding to feedback and trying to make it better.
In that time period, shortly after launching and listening to the feedback, I had more questions and did a few deeper investigations on privacy and search privacy.
I have a graduate degree in technology policy, so these internet policy issues interested me. It piqued my interest when people were asking about the privacy policy.
Here’s one example. In 2006, AOL released a data set of search queries that was supposedly anonymous, but they were tied together via a unique identifier and people were able to de-anonymize a bunch of people in it.
In 2009, Netflix had this Netflix Prize where they also put out a data set that they claimed was anonymous, but it had session data and people were able to de-anonymize the data.
And then there were some high-profile cases, Yahoo among them. It was not exactly search related – that was more email – but it was about storing information on people and then handing that data over to China. The person responsible was jailed, culminating in a hearing where Yahoo went up in front of the US Congress.
This investigation made me realize a couple of things.
One was that private information could have real-world consequences to people like dissidents.
Ten years ago, policy people were thinking about things that are starting to come true now.
What if health insurers got access to your searches and could charge you different prices? What if retailers knew your search history and could charge you different prices? Both of those things are happening now. What if governments across the world, not just China, were starting to use this information against you? I realized that it could be harmful to people.
The second revelation was, it wasn’t necessary to track searches in order to deliver good search results or ads. The business itself is uniquely positioned to not store any information because you type in your intent and you can give ads based on that intent. The search results generally are based solely on the keywords you type, and you can still derive signals anonymously. Just see in aggregate what people click on, nothing to do with the user session.
And then the third thing I realized was that any kind of user session can be de-anonymized. There’s no such thing as anonymous session data, as far as I’m concerned. As soon as you have so-called anonymous session data, you can find something to pair it with that de-anonymizes it.
So if you take all those things together, it seemed like you could run a search engine that would be a better user experience for the user.
So I made that decision in 2009.
In early 2010, I publicly announced that’s what we did. But at that point, it wasn’t that much of a motivator in terms of attracting people to the site. It was still pretty much an edge concept, because the harms were not very well known.
There was interest in a small sub-set of the community, but we were still attracting people more for those other reasons, like instant answers, customization, less spam.
All of those dwarfed the privacy reason for a while.
Damian: And then you launched as DuckDuckGo?
Gabriel: Then we launched as DuckDuckGo.
Over time, it became clear that there were very few things that allow you to compete against Google. It has to be something that they can’t copy technically, which is easy for them. A lot of the features that we had originally used to differentiate ourselves were eventually copied – spam and instant answers.
It has to be for another reason, a legal, business reason. So one of the remaining things that differentiated us was privacy. For structural reasons, they have gone – and continue to go – in a completely ‘get as much data as you can’ direction.
And as they march in that direction, our differentiation becomes more clear. And as the world has gone in that direction, it becomes more needed.
So then, in 2013 is when the real step function happened, with the NSA revelations and the recognition that actually some of the things we were talking about and the reason why we don’t track were true.
There was this backdoor gateway to searches in Google, in particular, that the government has, and so that was the one sticking point for government surveillance.
Since 2013, the corporate surveillance side has increased, increased and increased.
That had its big sticking point in 2018 with Cambridge Analytica.
We were seeing it way before, though. Even back in 2012, Google changed their privacy policy to allow sharing data between its services, which is now why people are up in arms – correctly – about Facebook.
Google were anticipating that earlier because they had bought DoubleClick and then started finding these things; they wanted to store data, and then over time, almost every year there’s been a walking back of any kind of Chinese wall between different things in their privacy policy.
The last one was, I think, in 2016, where they finally said, basically: ‘We’re building a big profile of you and we’re using it across all of our advertising networks the same way.’
Your searches will follow you around. That has happened more and more, along with various harms like the filter bubble and manipulations through ads. All those different things have become more and more harmful. We stand as an alternative.
Damian: Do people give a shit? On Amazon, the convenience of getting a Swiffer mop within two hours is a killer. Convenience wins, right?
Gabriel: Yeah.
Damian: This was really the most interesting revelation for me: convenience is such a drug that people seem to talk a good game when it comes to privacy, but I didn’t see that many people do anything about it. Your numbers would suggest otherwise.
Gabriel: There are a couple of ways to look at this. We’re more than just a search engine, but let’s discuss search for just a second.
Let’s say we were literally exactly the same – to the extent that there was no sacrifice whatsoever in switching, even in terms of design. Then it becomes a no-brainer. I think people really do want more privacy if they can get it for free. Right?
And so, the closer you get to that line of ‘you’re just as convenient, you’re just as good, potentially even better in some cases’ then the more people switch.
Because people do care. They do want privacy. The question is how much do they want it? How much are they willing to give up in different situations?
We found as our product gets better and better, and it’s easier and easier to switch, then we have more adoption.
Then there is the second question. For different people, in different places, there are all sorts of different privacy harms. They care about privacy to different degrees. People are willing to go different amounts of sacrifice. And so we find that as the awareness of the harms increase, people are willing to do a bit more, right?
That’s what we’ve seen over the past few years. The harms have increased. There are more and more creepy ads. There are more and more data breaches. There’s more and more identity theft.
Those three things drive a lot of interest.
People are seeing those harms a lot more. And it’s been easier to switch. So you see adoption grow as a result.
There’s another piece. We realize privacy is really complicated for people to understand because it’s a really complicated concept. There are all these different harms, all these technologies you have to use to help yourself.
We see our role as simplifying – both in terms of education but also in delivering the one tool that can help you reduce your footprint online. And so our latest iteration of our apps and extensions have all the privacy essentials in them that you need.
You use it to search, but it’s also tracker blocking, encryption, all in one package. And what we can say to people is: ‘It’s available on every browser and device. All you have to do is find DuckDuckGo, and you’ve kind of done the 80/20, and you’re in this more safe environment.’
So, from people’s perspective, there’s this big bad internet where all these companies are trying to exploit them, and if you adopt DuckDuckGo, you’re not having all those sacrifices, just changing your browser and search engine, and it’s a nice, safer world. That trade-off is easier to make.
That said, I think if the sacrifice was really great, then I think you’d be right. If we were charging people ten bucks a month, we would have a lot less users.
Or, if you couldn’t get any search results. If it was really, really worse, then it wouldn’t work either.
Damian: Do you have profiles of your users?
Gabriel: So this is quite interesting. We have a blog, it’s spreadprivacy.com.
There’s the research side of it where we’ve continually done mainstream research, surveys on privacy, trying to figure out our target market: who says they’re concerned. But beyond that, who actually acts on it. Who’s actually adopting DuckDuckGo or other privacy tools?
The interesting part is that it cuts across all demographics. There’s no signal in that, in normal demographic terms. It cuts across political parties, ages, income, education. We’re finding essentially representative groups of all.
Damian: And country specific?
Gabriel: To a degree. North America, Europe and Australia, New Zealand – we have generally similar market share adoption. It’s not like there’s a huge spike in Europe versus the US or anywhere else.
Damian: Where are your servers or who’s hosting them?
Gabriel: AWS. We now are in five different regions but we can be in more. We just haven’t done it.
Damian: And you don’t have any concerns in having your data hosted on Amazon Web Services?
Gabriel: So we’ve constructed it so that we are in control of all the servers, and we basically don’t use any of their tools that involve handling encryption keys.
This is a nuanced topic, but I think it’s actually more secure than hosting on a random hosting platform.
You have to host somewhere, right? And so if you’re in a random hardware that isn’t yours there’s a whole bunch of other physical security concerns that one has that are harder to control for than, I think, in the virtual server environment.
In any case, the short answer is we encrypt everything to our boxes and only we see it. We also don’t log anything in terms of personal information. So it’s only sending on the wire to us encrypted and then sending back encrypted and then we don’t store anything.
Damian: When I’ve explained DuckDuckGo, I’ve struggled even to explain how something like DuckDuckGo can serve up Google search and it not be part of Google. Could you explain for my mum? How is it safer or simpler to use you guys than it is to just use Chrome?
Gabriel: There are multiple ways to adopt us. You can use Chrome and install the DuckDuckGo browser extension.
What that will do is twofold. When you surf around the web, Google is tracking you everywhere you go, pretty much. They’re on like 75 per cent of websites, lurking in the background.
What our extension does is blocks their trackers across the web – as well as Facebook’s and all sorts of other data brokers, bad actors.
When you search, there are more search engines than Google. So they can be delivered by somebody else. We provide a private search engine so that when you do a search, we can serve results to you anonymously. So when you search DuckDuckGo it’s like you’ve never been there before. Whereas Google saves all your search history and then uses that to target ads to you across the web.
When you use DuckDuckGo your search is anonymous. Then when you click on those sites we block Google, as well as trackers.
Then when you click on those sites we also block Google on those sites.
So you’re experiencing the web, kind of Google free, essentially.
The most practical effect is that you immediately stop seeing those ads that follow you around everywhere. That’s a sign that people aren’t following you any more.
Damian: But couldn’t you just use private browsing mode on Chrome? Isn’t it the same?
Gabriel: It is not the same. That is a big myth that I would love to dispel.
Private browsing mode: all it does is scrub the data from your computer when you’re done with private browsing mode.
It does not prevent anyone on the other side of the connection from tracking you. And so when you go to a website in private browsing mode, that website can still see your computer IP address and everything about your computer. They can associate that with a profile on their side. So even though your history’s deleted on your side, they still have it on their side.
So Google can still be tracking across the web even in private browsing.
Damian: So how do we educate people more?
What experience have you had in educating people that works?
Gabriel: I’d say two broad areas.
We use the blog as a platform to kind of get wider mainstream press on the issues. So we’ve been doing original research or original commentary that then gets picked up and so I think that has worked pretty well.
We’ve been writing long-form Quora answers and other blog posts.
They really go in depth and explain these things. We’ve been advertising those to get people on them at points of interest. That’s actually worked a lot.
It’s a good selling point. With WeTransfer, my sense, just looking at it, is you could do more in that messaging inside the product, somehow, to educate.
What we found is, we spend a lot of time writing these long-form things in a really mainstream, readable way.
In your case, to the audience of creatives, one day each creative will be in the moment of wanting to read this thing. Then they’ll finally read it and you’ll slowly be educating them, over time.
It won’t magically work the first day you put it up. But it has worked, over time.
For example, on the DuckDuckGo footer, you just search and you look at the bottom. We have these cards. Some small percentage of people click every day, but over a long period of time, say the next two years, our assessment is that most will eventually click on it. You know? And then they’ll maybe read it.
Damian: Are you optimistic about the future? Do you think that we’re going to get to a place where people are much more educated around privacy?
Gabriel: I think so. I’m pretty optimistic about it after last year.
Damian: What about Google? Where are they gonna go? What’s gonna happen to them and Facebook?
Gabriel: I’m not optimistic that they’re going to implode or change their practices very much, but I’m optimistic that people are getting awareness, making different choices.
That is translating into some kind of privacy legislation in different places worldwide, including the US, where people basically laughed at you if you mentioned it a year ago. You know?
So that has drastically changed. That’s a good environment to be optimistic about. And generally, what you find here is that it all follows from the will of the people.
People say: ‘Oh that’ll never happen.’ But as soon as you start getting polling numbers, people start voting on an issue, everything can change quickly.
That started happening on privacy – at least, that’s the test this year. If nothing really happens after this year then I might be more pessimistic next year.
•••
In any book about trust, it’s important to speak to those working in industries considered untrustworthy. Andrei Cherny is the CEO and co-founder of Aspiration.com, a rare bright spot in the banking world. Much like the real-estate industry, banking is dogged with a legacy of corruption, overcharging, contempt for customers and unfairness. For many, the day Lehman Brothers closed down is now the distant past. But the financial industry doesn’t seem to have learned anything in the years that have passed. Banks have once again started trading credit default swaps. Student loan defaults have increased. The industry has found inevitable loopholes to escape the consequences. Where there is money, big money, there will always be a way.
A bank that treats its customers fairly is hard to find. A bank that gives back to its community is a nice notion and probably exists alongside Sasquatches in some alternative dimension. We can make jokes. We can grumble, but it’s important to remind ourselves of the bitter truth. We need the banking industry. But we need more banks and financial institutions that behave like Aspiration. Ultimately, it’s these lenders who will set the goals and levels of ambition for business – in particular tech business. And today, every business is a tech business.
Unsurprisingly, the issue of trust came up right away with Andrei. It seems to dictate much of what they do.
Andrei: Internally, when we’re having a meeting, a discussion, and just wrestling with a problem, our motto is ‘solve for trust’. When you’re looking at banking and the financial industry, you’re looking at an industry where 8 per cent of Americans say they trust their financial institution. And it always seems to me that if you’re not first and foremost solving for that problem, you’re missing the 800-pound gorilla in the room.
It’s great to have lower fees, and better technology, and cool features. All of those things are necessary, but they’re not speaking to the central core challenge that people have, especially when it comes to the financial industry.
The banking experience used to be a community experience, a local experience.
There was the experience of having a person that you knew, who you saw at your kids’ Little League game or supermarket.
Then there was a wave of conglomeration. That local experience has now disappeared. And so there’s a space for using technology, not to replace human interaction, but to scale human interaction.
You had a financial industry that made a pivot in how it made money – to really focus on fees as the main driver of interaction, and fees around financial illness as opposed to financial health.
The big banks make money when things go wrong for you. When you have an overdraft fee, when you have a late fee, an out-of-network ATM fee, when you have monthly service fees because you don’t have enough deposits in your account, and so on.
That means there’s a fundamental misalignment of interests between people and their financial institution.
We wanted to solve for that, and we also understood that there’s a misalignment of values. People’s expectations of the places they’re doing business with, in this day and age, have changed. They’re looking to those businesses to have a conscience and a sense of ethics and values. And that certainly wasn’t the case in the financial industry.
Damian: I’m trying to find companies that are ethically minded, or values driven, or have strongly ingrained CSR [corporate social responsibility] approach within the business. It’s exceptionally hard to find it within anything other than retail: Patagonia, TOMS, Ben & Jerry’s, those guys have been doing it for between ten and twenty years. We’re a Dutch company, so banks like Triodos have been around quite a while, and are quite well known in the Netherlands, but not well known really outside of it.
But in the tech and the finance world, there are very few companies that have built into their DNA an ethical approach to business. I can barely think of any.
Andrei: For us, the two were always linked.
Previously, I had just spent a number of years in the public policy space and worked in the White House and the Clinton administration, and worked with Elizabeth Warren on the idea for what became the Consumer Financial Protection Bureau. I was a financial fraud prosecutor, and so on, doing consulting, and some of it for big banks and large institutions. I saw these very large institutions who were trying to bolt on a corporate social responsibility program.
People saw right through that.
We said we didn’t want to be a business that was making money in ways that were fundamentally at odds with the best interests and values of our consumers – and then do some nice things on the side. We wanted those two to be inextricably linked. We would only make money to the extent that we were doing the right thing in the right way for our customers.
This led to the ‘pay what is fair’ fee structure, where we said we were going to let the customer choose the fee.
It’s not a free service by any means. There is a fee. As opposed to us setting the fee, we want the customer to pick the fee, and if they think we deserve to be paid zero, then we deserve to be paid zero, and we need to be doing a better job. But then it’s up to us to deliver for that customer – whether it’s in the service we provide, or the values – in such a way that they choose to pay us even though they don’t have to.
We were going to believe that our customers were going to treat us with the same kind of honesty and integrity that they’re expecting us to show them.
That seemed natural. We really wanted to do things differently and build a strong charitable commitment into it as well.
It’s been three years now.
Damian: Can you demonstrate that it’s working?
Andrei: Which part? The business?
Damian: Here’s what we’re competing with: a huge chunk of the tech industry is very financially driven, so a lot of the employees are simply going for the highest salary, the best perks, and the best options package, which leads to a phenomenal amount of churn.
I’m not going to compete with Google and Snapchat and everybody else, because I don’t believe in competing on purely financial terms.
What we’re offering is a different mindset. But I haven’t yet found a way of putting it into any sort of numerical system that actually can justify how it’s working. Have you managed to achieve something like that?
Andrei: We see this working in terms of customer growth. We’re nearly at two million users at this point, and growing very rapidly. We see it in our ring of social media engagement.
You see the vast majority of customers choosing to pay, even though they don’t have to.
That really speaks to the depth of that relationship we’ve been able to forge. We now have over a hundred employees. We’ve always insisted that we only hire people who are very mission driven.
I don’t have a numerical value I can assign, but I certainly feel what we’re going to offer our employees is something different.
If you come into our offices we have IKEA furniture for our commons table. There isn’t art on the wall. We don’t have free lunches for employees, or other perks like that. But we also make sure our customers never have to pay an ATM fee, and we give 10 per cent of everything we make towards charitable giving. And we say to our employees, if you like those other things there’s absolutely nothing wrong with that. There are plenty of other companies you can go work at.
We provide something different, and we’ve built a team and a culture around that.
Damian: Who selects the charities? Do you have a CSR department or is everybody involved?
Andrei: People are involved in different ways. What we’ve at least been historically doing is working primarily with a single charity, which is Accion, around charitable micro loans.
Accion USA network is the largest charitable provider of micro loans in the United States. So they’re giving loans to low-income individuals who are looking to make a change in their life, and so that’s been our charitable partner and they’ll tell us stories around that.
Damian: You’ve either been really lucky with your investors or you’ve chosen them very strategically. Not every investor would be interested in a business like yours, right?
Andrei: We have been lucky. We’ve chosen strategically and, to be honest, we’ve had an endless number of doors close in our face as well, along the way.
Damian: I’m glad to hear that it wasn’t just me.
Andrei: Look, we started a business – my co-founder and I – saying we were going to be a financial business focused on everyday people and not the very wealthy. We were going to let people pay us whatever they want, including zero. We were going to give 10 per cent of what we earn towards charitable giving. As you can imagine, there were initially some very short conversations.
It’s interesting. The first major investor who really believed in us was Jeff Skoll, who was president of eBay, and created the Skoll Foundation. I remember sitting down with him and describing what we were doing, and he said: ‘Guys, I think this is going to work.’
I said: ‘Well, maybe you didn’t hear me. Here’s what we’re doing.’ And he said: ‘Nah, nah. I think it’s going to work.’
This speaks directly to what you’re talking about. He said: ‘Every ten years or so there is a company or companies that come along that make a big bet on trust.’ He said that is what made eBay successful, twenty-plus years ago.
In the early days of the internet, it was the Wild West, and all of their competitors had these complicated escrow models.
You’re just going to have to believe that this person you’ve never met before, that you’ve paid money to over the internet, is actually going to send you that baseball card, or whatever else it is. And similarly, if you said ten years ago that there would be big businesses based on getting into a stranger’s car, or staying in a stranger’s bedroom in a foreign country, you know, people would have said that’s crazy. And certainly you’ve had the Airbnbs, Ubers, Lyfts, and so on, that have come along. And he said similarly, you’re taking this big bet on trust. It’s going to pay off for you.
•••
I wanted to take a break from the story of my company to ask Jimmy Wales, co-founder of Wikipedia, what we’re going to do. By that I mean the greater ‘us’, the general public. What are we going to do about the internet of the future, because, as I told him, I fundamentally believe it’s down to you and me, as in ‘those people in the street’. I’ve discussed what a responsible tech company like mine could do. It’s not down to business people to shape what the internet looks like. The user has this idea that they have no control over what’s happening. I truly believe the reverse is true. All the power is in the hands of the user. They just feel powerless.
Some people who have been very much involved with the internet from the very beginning, or from the early days, are still absolutely 100 per cent smitten with what the internet can do and completely unperturbed by Big Data and the potential consequences for the future. And there are those who are absolutely paranoid and living in fear every day, because their five-star rating is basically a make or break for them to receive the next job offer.
I wanted to talk to Jimmy about how he built trust and what – obviously – the future is going to look like.
Damian: How old is Wikipedia?
Jimmy: Wikipedia is now seventeen.
It’s funny because I never really think in these terms about Wikipedia. I never think of a market for Wikipedia. It’s just not the way we ever approached it, which is interesting, I think, in and of itself. In a sense, we are an alcove of early internet culture. But I don’t think that meant that we thought that Wikipedia was only for hobbyists because already, by the time Wikipedia started in 2001, the internet was already quite big.
We had already been through most of the dotcom boom, and Google existed. Facebook didn’t yet, but lots of things that we know about were already familiar and existing. I didn’t think of it as a hobbyist market, if I thought of it as a market at all.
Internet on your phone was, if at all, very hard to use and basically didn’t work, like those pre-smart phones. So it was very much desktop: laptops, and real desktop computers.
In that sense, obviously, it was very different than the environment we look at today. Maybe slightly more than 50 per cent of our traffic is by mobile. So that’s different in a lot of ways, including what does it mean to participate at Wikipedia? Obviously, the vast majority of people who are editing Wikipedia are on a desktop. Editing from a little device is not particularly comfortable. People do it, but it’s not easy. Which means that there are loads and loads of readers who probably won’t convert very easily to becoming authors, potentially because they’re readers on their devices. That’s an interesting kind of dynamic that’s going on.
Damian: So in and of itself, it’s pretty revolutionary that people decided to build something that didn’t have a business model behind it.
You haven’t, for whatever reason, felt pressured to put one in and do the normal premium service, the pro-version, anything else?
Jimmy: I would think about that in two ways. In the early days, the very early days, it was not unusual, but only because it was the dotcom boom, when I first set up Wikipedia. And at that time, it was really a mentality of ‘Don’t worry about your business model right now. Just build something.’ And so it was pretty common. I mean, obviously, we have some spectacular flame-out failures: services that became quite popular, but literally had no coherent business concept whatsoever. At that time, I remember, Webvan was a famous one that lost, I think, $800 million from doing online delivery of groceries, which, of course, is a growing and big business now, but back then it was just they couldn’t make it work. But that was okay. At the moment, they were like, ‘Let’s just get big, and we’ll figure it out.’
And then we had the dotcom crash, and so in many ways, Wikipedia is truly a child of the dotcom crash because, as it began to grow, there was no hope of raising money.
Silicon Valley wasn’t investing in anything because everything was a disaster beyond all measure. And that was, in some ways, a blessing for us because a lot of the innovations of Wikipedia came about because we literally had no money. There was no way to solve problems by throwing money at them, which means we had to solve them, but maybe solve them in a more innovative way.
Just a simple thing would be, you would imagine, as a site starts to grow, if we had 10 million, 20 million in capital, we would have said: ‘Oh, right. Well, so we’re starting to see this problem or that problem. We need to hire community managers and moderators, and pay staff to block people and to enforce the rules and so forth.’ And we would have gone right down the model of YouTube and that sort of thing, where the community actually has no power whatsoever. And the moderators are staff members. And that would’ve been very, very different, and I think not nearly as successful, nor innovative. But because we didn’t have any money, we had to figure it out ourselves – or the community would. We had a lot of the checks and balances that you wouldn’t have developed if you had just paid people. You have to sort of say: ‘Well, who are the moderators going to be?’ It begins to look a lot more like a democracy in the sense of a good municipal government. You need people to enforce the rules, but they have to be accountable. And so the admins, all their actions have to be transparent, and they have to be able to lose their admin rights. And who’s going to decide that? And you have to have procedures and processes and things in the community. So we built up all of those things.
How admins are elected by the community, and then how they can lose their admin rights. Things like the arbitration committee, which enforces higher level – kind of like the Supreme Court of Wikipedia. It’s a whole set of things that we had to evolve, but they were evolved in light of saying: ‘We’re just a volunteer community.’ And so that also gave us incredible intellectual independence.
Wikipedia from the very early days didn’t cost a lot to produce because it wasn’t done by staff. It was done by community members.
I decided: ‘Okay. Look. We are a non-profit. We need to ask for money. We need to do a fundraiser.’ So we did. We announced a fundraiser, and I was hoping to get $20,000 in a month’s time, and we had nearly $30,000 within about two weeks’ time. So it was a huge success. But that was the first time I really thought: ‘Well, just asking people to support Wikipedia might actually be a model that works.’ And it has proven to be so.
One of the things a lot of people don’t know, although we’re super transparent about everything, is if you look at the finances of Wikipedia, it’s kinda getting to … We’ll call it embarrassingly profitable, in the sense that we run the organization in a very financially conservative way. And every year, we spend less than we bring in, and our surpluses have been over 20 million, 15 million. I don’t know what it’s going to be this year, but we had our best revenue year. We just broke 100 million in revenue. I’m sure we didn’t spend even 80.
So as a financial model, it’s proven to be surprisingly successful, but it’s a charity. So obviously all we do with the money is we just build our reserves for the future and spend it on charitable projects, but it’s kind of interesting.
Damian: Is it fair to say that you discovered this by accident?
Jimmy: Yeah. Sort of.
There’s a lot of luck involved in anything like this. So of course, there were a hundred decisions a day, every day, not just that I made, but that the community made. It sort of [came about] by accident, but also a lot of trial and error, a lot of hard thinking about the problems and figuring out how to do it. So we didn’t know everything a priori, but we had certain conditions a priori, and then we just stuck to those conditions and solved the detail problems as they came up.
Damian: So at any stage, you’d never actually taken any investor capital?
Jimmy: No, zero. Zero. Yeah, Wikipedia is completely self-funded.
Damian: So had you taken investor capital at one moment, it would be a completely different story?
Jimmy: Oh, completely.
Damian: $200,000 from somebody, it would’ve changed everything?
Jimmy: Totally, totally. Because we would have ended up a for profit, with all the needs that go with that.
Damian: You’re a complete anomaly in this because you, as a developer, have taken a very empathetic approach to an online business. You could have changed it massively. This is perhaps not something I’d say publicly, but often, developers are not that empathetic. They’re not that community driven. In that sense you are unique.
You could argue that, had Facebook struggled in the way that you struggled, and launched at the same time, Facebook could well be in the same place as Wikipedia today.
Jimmy: It’s at least conceivable. Certainly, when we think about the business models of, for example, Facebook, the business models are always the product. Not in every way, but in a lot of ways.
That’s kind of interesting to think about something that’s so fundamental to society in the social sense, how everyone’s using Facebook and sharing photos and things like that. If you think the business model drives the service, and it’s Amazon, that’s a lot less weird in some way because it’s like, yes, Amazon is an e-commerce site that wants to sell you stuff, so the whole business model is about making it easy for you to buy stuff. Make sure the prices are right.
Everything about their business model is consistent with what you’d expect, in an obvious way. Everybody understands it. Whereas with Facebook, there’s a lot about their business model that is … people are aware … the famous line when Mark was speaking to Congress. When he was sort of baffled. ‘How do you make money anyways?’ He’s like: ‘We run ads.’ Almost everybody except the Senator understands that, but the deeper question is: ‘What does that mean in terms of what do they want to drive you to do on the platform? What is the information they want to elicit from you?’
These people are just now beginning to grapple with [the realization that] this model doesn’t drive Facebook to make the most wonderful sharing platform, it drives them to get your data. Whereas the business model of Amazon does drive them towards making the most wonderful shopping experience, so it’s more aligned.
One of the things that I really talk a lot about with respect to Wikipedia and its business model is the fact that it’s a charity. When I say ‘this model’ in this context, I mean people just donate money because they love Wikipedia and they think it should exist.
If we were a non-profit and we had an advertising business model, then the DNA of the organization would inherently change because organizations do follow the money. That’s an inexorable fact. Company culture can resist certain things, but at the end of the day they follow the money.
Right now, the foundation staff care as much about the next million readers in sub-Saharan Africa as they do the next million readers in California because the business model is about being a charity and about making people love us and feel like: ‘Wikipedia is the one thing in all these troubling times that I think is basically decent in the world, and I just think I should chip in and support what they’re doing.’ Which should drive the staff to have a very different set of incentives.
If we were ad focused then I could tell you we would definitely be having an internal meeting looking at revenue. If only people would search more for cheap hotels in Las Vegas rather than Elizabethan poetry, our revenue would jump, so then we’d think about ‘let’s appease them with clickbait-y headlines and that will get engagement up and da, da, da’ and we just don’t think in those ways at all. I think it’s really interesting to think about how, in some subtle ways, business models drive product in ways that are maybe not always healthy.
Damian: Do you use Facebook personally?
Jimmy: Yeah. I do. I’m a very ordinary consumer of the internet. I use all the usual stuff that everybody uses and I actually like Facebook, so obviously you can see I have great concerns about where their business model drives them. It’s a great product. Say what you will about Facebook, it totally works for what most people are trying to use it for. So there is that.
Damian: So you’re not concerned?
Jimmy: Sometimes I worry. I worry not in the abstract, I just worry because I’m a non-person on the internet. I’m super careful about all my privacy settings and security, just because I am. In general, I’m like most people. In the last couple of years, the crisis of conscience that happened with the public and the internet companies is kind of the realization that there are side effects. I love programmatic advertising. I love that no matter where I go on the internet, I get ads that are relevant to me. I think that’s fantastic. I see a lot of win-win-win in the sense that, if I’m seeing ads that are relevant, that means that the publishers running those ads are going to get paid more because it’s not just punch the monkey and all the stupid Tammy ads.
That’s great for them, it’s great for advertisers because they’re partnering with more people. It has the relevance. It’s all great. Then you begin to understand: ‘On the side, we’ve got all this weird stuff going on with Cambridge Analytica slicing and dicing for political purposes.’
For me, one of the most disturbing things about that is, in the old days, politicians in order to win had to have a single consistent message, then reach across the middle. That got enough people on board with it. Whereas now it’s beginning to feel like they can run campaigns that completely contradict each other. They sell completely different messages to different audiences, and a lot of it’s just voter suppression.
That’s disturbing, and you realize: ‘All that data is out there, and it can be used for more than just making sure that I see boat ads.’ That’s what I always say, because I like boats. It can be used for all kinds of questionable purposes. That’s a dark side effect that I think people are really trying to grapple with.
Damian: You guys are such a beacon of hope for so much about the internet, because there’s so much discussion around should the internet be regulated or not? If you look at Wikipedia, Wikipedia has proven that it can actually self-regulate. How does regulation fit in?
Jimmy: I’m no fan of regulation. I’m a First Amendment fan and I’m an entrepreneur and all of that classic stuff. I don’t think governments generally do a good job even understanding the internet, much less trying to pass nuanced regulation. So I’m already sceptical. If you’re proposing some kind of online regulation that will impact the way people are using the internet, freedom of expression and that sort of thing, you have to say: ‘Is this the least intrusive way to achieve a legitimate interest, or are we just splashing around at the behest of the music industry?’ or whatever it might be.
There’s that. But I’m also very sympathetic. A lot of those problems just don’t come up for Wikipedia that do come up for other companies. I just saw there’s a big outcry. The other day, Mark Zuckerberg said something along the lines of: ‘We won’t take down or block people who are engaging in denial of the Holocaust.’ People were outraged, because that’s really stupid behaviour, people denying the Holocaust, it’s hateful – and Mark Zuckerberg is Jewish, even.
But I say: ‘You know what? At Wikipedia, we’ve never allowed Holocaust deniers and we’ve never allowed loads of things, but also we’re not a community free speech platform. We’ve always said we’re here to build an encyclopedia, we’re not a platform for people to post all of their opinions. You’ve gotta have a reliable source. We have an article about Holocaust Denial, but it’s quite neutral and, in some ways, would be quite negative because it’s historically complete nonsense.
That means we don’t have to deal with it. When Twitter struggles with how do you distinguish between legitimate political commentary and saying Donald Trump is an idiot? Or saying Obama’s an idiot? Not about Donald Trump. Although he is, in fact, a massive idiot … So saying a famous politician is a bad person, insulting them, being rude about them, that’s just part of the discourse. That’s part of the world, and we don’t want to see that kind of thing banned. At the same time, if it’s a group of teenagers going in together to bully someone from their class, being insulting and telling lies and getting things wrong and spreading rumours, then we’re like: ‘Actually, Twitter, why don’t you do something about that? It’s terrible. Why are you letting this kind of hateful behaviour go on?’
I’m sympathetic. It’s a really hard problem for those kinds of platforms. To imagine that politicians can step in and define a magical set of rules that will make the world right is really naive. That really underestimates how complicated that problem is.
When we say, yeah, Wikipedia is a beacon of how you can self-regulate. Yes. Lovely, and I agree. And I think it’s a very kind thing to say. But I also would say we’ve chosen a much easier problem, in a sense. Which is to say, if you go on to Wikipedia and you write Donald Trump is the worst president in history and I really hate him and you want to put that into the Wikipedia entry, our community’s going to go: ‘That’s not an encyclopedia, that’s just your opinion. Do you have a source?’ It’s a much easier problem. You deal with hayseeds by going: ‘That’s not what we do here. We don’t care what your opinion is. We’re documenting the facts of the world.’
Damian: Your governance is, and has been, strong from day one. I don’t think that Twitter and Facebook had that governance in place from day one.
Jimmy: What’s interesting about that is on a personal level, when I think about my values and things that I think are important.
In terms of a real authority and power, I’m one of ten members of the board. That’s the extent of my power, plus my public presence, which gives me more influence than a lot of other board members in my position. But the truth is, control is not something I actually have … there’s been a few little things here and there where I think I would have decided differently from staff. And I think this is horrible, or it’s like: ‘Oh, okay. Well, I would have done that differently, but actually I see the other side and it’s fine.’ So I’ve been very lucky that the organization hasn’t sort of struck off in a bizarre direction.
And I do think that is about the full scale of the way it all works with the community, and the structures within the community, and all of the things that we designed very early on being fairly robust over time.
The Wikipedia community, I would say, is slightly left of centre, but they’re not going to call people white supremacists because that’s quite an extreme thing to say. Somebody put it in and somebody removed it very quickly. It was not a huge deal. It’s just the noise of Wikipedia. You could be critical. Why does Wikipedia allow that kind of noise? Okay, that’s a valid question, and I’ve got a good answer for it, but it’s not like suddenly I’m fighting against a community who’s gone radically left wing or radically right wing. We haven’t been taken over by alt-right trolls or anything like that. That’s good, and it’s healthy that the system is reasonably robust, and obviously there’s no guarantee that it will always stay so, but there’s a lot of really sensible people, and so you feel comfortable. I sleep at night.
And I think, if somebody says: ‘Oh, it’s terrible about your Wikimedia.’
‘Oh, what do you mean?’
‘I hear that it’s been taken over by total right wingers.’
I’m like: ‘Oh, dear!’ No. I feel that’s not true, actually. And it isn’t true.
Damian: I take a metaphor from the physical world. I think America is the epicentre of the boomtown. And what we’ve seen with the internet is this rise of pioneers, in the early days, who came in and set foundations, created community, failed, moved on slightly, built something else.
And some of the empires that we’re seeing today are Facebook, Wikipedia, those sort of places. But I think we’re going to go through a transformation in the cityscape of the internet, where we’re going to see a lot more of the sort of strip malls disappearing as advertising revenue disappears. And more and more, I think we’ll be moving to different revenue models. Subscriptions will be one part of it, but I don’t know that I want to have a hundred subscriptions. What do you think? What do you think in twenty years from now is going to be the cityscape of the internet? How do you think it would look in terms of business models or community?
Jimmy: I think it’s really hard to predict. There’s a lot of different things that might happen.
I was on the internet which was academic and small. I knew there was America Online and Prodigy and CompuServe. And there was a belief, which I think was not entirely invalid, that whoever got biggest fast enough would win, and that it was going to be a monopoly business. Why would you sign up to AOL versus CompuServe? Well, you sign up for AOL because all your friends are on AOL, and that’s your starting environment. There is no internet, in a sense. It’s just that.
We dodged a bullet there. I don’t think it was absolutely guaranteed that the internet was going to become the dominant online access, versus a company that managed a whole big process. Now maybe I’m wrong, maybe it was too hard for one company to scale. But I can envision if a company were smart enough, if AOL were smart enough to forego revenue, and encourage other companies, they’d say to whoever the predecessor of Amazon would have been: ‘Come and set up your store here, and we’ll give you all the tools. And we’re going to lose massive money doing it.’ It might have ended up that we were all on AOL, and a single company would have controlled the whole thing.
But we didn’t end up in that world. And whether that was ever a real alternative or not, I don’t know. I actually think it was possible that we could have ended up in that equilibrium.
And so similarly today, when I look at various changes in the landscape, I do think there are interesting ways that we might end up in some kind of bad corner solution.
And then there’s a lot of interesting ways where we end up still with a diverse, open internet that is the dominant platform, and it is a place that is neutral and everybody can join and so on.
I generally tend to agree with you. But I also have some concern that I might just be predicting the future based on what I hope is true.
See, if what I hope is true comes true, and I do think a large part of it will … I do think we’ll move away from advertising-only business models. I don’t think they’re healthy for everything. I mean, for some things, fine, whatever. But I’ve been doing a lot around journalism. I’ve got my new project which is premiering, which is about communities and journalism and so on.
But one of my big concerns about journalism is that the advertising-only business model has been incredibly destructive to quality journalism, and the reason is quite obvious. The incentive that it puts in front of publishers in an era of programmatic advertising is just clickbait: as many eyeballs, as many pages as possible. Attract the right audience for five seconds and show them the high-value ad. It’s basically a race to the bottom, clickbait-y content. And then serious publishers struggle, because serious content costs a lot of money to produce.
Meaningful things that change people’s lives cost a lot of money to produce.
If The Guardian has a senior reporter working for three weeks on a story, and that story is important, and it moves the needle, and it gets a lot of attention … and if at the same time, a clickbait-y publisher publishes a listicle, as they call it, ‘23 Cat Photos, And Number 7 Will Make You Cry’ that some millennial intern made in three hours of an afternoon, and they’re clever and funny, and it makes the same amount of page views and the same amount of money, then The Guardian has to look at that and goes: ‘This business model is not working for us. We should just be hiring a lot of funny young kids to write fluffy content, because it’s cheap and it gets eyeballs.’ As opposed to what I’m really excited about is we’ve seen the New York Times, their subscriptions have gone from a million to three million in the last couple of years.
That, for me, is one of the most positive little indicators in society that, actually, people are willing to pay. And how do you sell subscriptions? You have to make a difference to people’s lives. So people don’t pay the New York Times because they saw something that made them click and they wasted fifteen minutes of their life clicking on ‘You Won’t Believe What This Child Star Looks Like Now’. They read a piece, and they thought: ‘Wow, that’s important. I need to support this. This should exist. I should pay for this.’ And the fact that we’re starting to see subscription models take off in journalism is really important.
In a lot of areas I think that can be very important. And a lot of the advertising-only stuff is going to end up being just … you won’t get quality there. You’ll just get noise, and I think that matters. I think people are going to be interested in saying yeah, actually, I want to pay for this because it’s worth paying for. Sometimes people talk about, in all the journalism, that the industry made a strategic mistake by giving away their stuff for free, and now people are used to it and they don’t want to pay for the news.
I think that’s actually wrong. I think that’s too simplistic a view. It was really hard to pay. So the more we have app models and easy payment methods, where people can just go yeah, this is worth it, I’m going to pay for it. And they don’t have to pull out their credit card, and they don’t have to type in all the digits and then worry about security and all that. The more people can just go yeah, that’s great, I want to buy that, out of an impulse purchase, then they will.
I agree with your overall view that people paying for content – people paying for services of various kinds – is likely to be a more dominant piece of landscape than it has been. And I think that’s a good thing.
•••