Gabriel Weinberg, Founder and CEO, DuckDuckGo
Taking on Google
Gabriel Weinberg has made it his mission to protect internet privacy, and his scrappy Google competitor DuckDuckGo is leading the charge.
Everything we do online is tracked. Searching, browsing, shopping, even navigating.
Most of us have grown accustomed to this. Although we may acknowledge now and then that it makes us uncomfortable, we haven’t changed our habits. Maybe we’ve become too reliant on our preferred online services for basic day-to-day tasks. Or maybe we don’t even know how we’d change our ways in the first place.
This is a problem that Gabriel Weinberg has been helping people solve since 2008, when he first created privacy-focused search engine DuckDuckGo. While the company has been charging ahead ever since, Weinberg’s mission is on everyone’s mind these days. Concerns about internet privacy and data protection are at an all time high, following recurring scandals around tech companies leaving their users vulnerable.
While still far smaller than industry leader you-know-who, DuckDuckGo’s popularity is surging, thanks to its commitments to never collecting personal information or tracking your activity to sell to advertisers. The search engine offers other services like informing the user of what tracking is being blocked, and now has a mobile privacy browser and desktop plugins. Since incorporating in 2008, DuckDuckGo has grown to a global company of 63 employees.
As internet privacy has taken the spotlight, Weinberg’s been busy, writing and advocating for federal “do not track” legislation, and speaking up in the New York Times opinion page and other platforms. He’s also got a new book out that explores the power of mental models he’s relied upon during his career.
But Weinberg’s core mission remains: to make it a lot easier for you to use the internet without being creeped on.
The Start of DuckDuckGo
DuckDuckGo is a search engine, but it’s also an internet privacy company that’s out to help you protect yourself online.
“We like to say the internet shouldn’t feel so creepy, and protecting your information should be as easy as closing the blinds,” Weinberg says.
DuckDuckGo offers a variety of tools to help consumers achieve this privacy—and feel confident about it. The company started as a search engine and has since expanded to offer a browser for iOS and Android, along with extensions for desktop browsers.
The company is 10 years in the making, but it wasn’t Weinberg’s first internet startup. After graduating from MIT in 2003, he created educational software that supported student achievement by using the internet to connect parents and teachers. Unfortunately, the software was developed about 15 years too early, and it fell flat.
Next, Weinberg started a pre-Facebook social network that helped people find old friends and classmates. That fell apart in 2006, and DuckDuckGo followed soon after in 2007. The company incorporated in 2008 and officially launched at the end of that year.
So, how has DuckDuckGo competed with giants like Google for over 10 years and lived to tell the tale? It’s a modern day David and Goliath story, although in this case David is growing bigger and stronger every day.
Weinberg attributes a lot of DuckDuckGo’s success to his team. As he often tells other entrepreneurs, “If you’re going to succeed, you’re going to need an amazing team around you. Work on crafting the values and mission to attract a team… to reach your ambition.”
The Power of Mental Models
Weinberg also credits much of his success to years of dedicated research on mental models, which he’s recently turned into a book with his wife Lauren McCann, Super Thinking: The Big Book of Mental Models.
“Mental models are concepts … be a better strategic thinker,” Weinberg says.
He encourages people to think of mental models like this: When you first learn arithmetic, you learn addition, then multiplication based on addition. If you didn’t advance to multiplication, you could still combine quantities using addition, but it would take you much, much longer.
Mental models operate the same way. “Once you know something, you can think in a higher-order way really quickly,” Weinberg says. On the other hand, if you didn’t have a mental model, you’d have to start from scratch, every time. It’d be more difficult and time-consuming to make good decisions, repeatedly.
When he started training the DuckDuckGo executive team, Weinberg realized a significant knowledge gap: His team didn’t recognize more than half the mental models he’d instructed them to learn and use.
That’s how Weinberg and McCann, a statistician, came up with Super Thinking—when he realized his current training method was inefficient and no other resource would suffice. Through their research, they also realized that many of the mental models were related.
Almost all 300 mental models in the book are existing concepts. Super Thinking simply collects and organizes them into main themes—nine, to be exact. The last two themes are called “Unlocking People’s Potential” and “Flex Your Market Power,” and the mental models in these chapters apply to leadership, management, and other business best practices.
“People are really different, and if you want to manage effectively, every person requires different characteristics,” he says. That’s why effective managers take the time to understand personality types and strengths, using questionnaires like Myers-Briggs and DiSC.
One notable mental model is called Joy’s Law, which tells us that all of the smartest people already work for someone else. “This means that you can’t corral smart people,” Weinberg says. “Instead, if you arrange people in just the right way and give them jobs that fit , they can reach extraordinary success both as teams and individuals.”
In their book, Weinberg and McCann explain that Joy’s Law also overlaps with mental models like 10x Teams and Resonant Frequency, the latter coming from physics. “A lot of mental models come from different disciplines,” Weinberg says. “The idea of mental models is to take a multidisciplinary approach—to take the best ideas from all different disciplines and combine them to use them for general strategic thinking.”
To understand Resonant Frequency at work, imagine an opera singer breaking a glass by hitting just the right note. “The same happens with people; hitting the right frequency and absorbing energy from the right role and jobs,” Weinberg says.
But just as a singer can’t break glass with every note they sing (it’s actually quite rare), a team can’t operate at a resonant frequency at all times. So, how do you know when you’ve reached it with your team? You constantly shift around your team and test to see which combination produces the best output.
That’s what Weinberg does at DuckDuckGo, where he doesn’t have a traditional management hierarchy. Instead, they’ve divided management responsibilities between positions and operate based on objectives and projects, and the team doesn’t hesitate to shuffle around when needed.
”We are constantly moving people around to fit what they’re most interested in and best suited for… to achieve these 10X Teams,” Weinberg said.
Gaining Traction for Growth
Weinberg previously authored Traction, which serves as a scientific experimentation approach to marketing. In it, he systematically lists 19 different channels that companies can use to gain traction with their audiences.
“My advice is to not leave anything out,” Weinberg says. “It’s often one of the unusual things might be the thing that actually works.”
Over the last 10 years, DuckDuckGo has found that different stages of growth have required different marketing channels. In the early stages of growth, the team used social media, content marketing, and PR, but these channels eventually saw diminishing returns. Since then, the team has transitioned to organic, viral growth and offline word-of-mouth marketing.
In the last nine months,DuckDuckGo has succeeded at brand marketing that has also raised market share. “I wrote some long-form articles on Quora on topics like why to use DuckDuckGo vs. Google, how tracking works, how to avoid it, and more,” Weinberg says. “These got such high engagement, more than anything else.”
Since publishing the posts, Weinberg and his team have been working to promote that content on platforms like Quora and Reddit. As one of the first native advertisers on Quora, their posts have been promoted to over 150 million people per month.
“It’s all about finding bigger audiences and putting content in front of them that’s compelling and native to the platform,” Weinberg says. The team has made just seven Quora answers and has been promoting them to 100 million people. It’s not cheap, but it’s working, he says.
What is DuckDuckGo’s traction like at this point? Since the DuckDuckGo engine doesn’t track unique users, they can only report on total searches, which is currently at over 1 billion each month. For a company competing in a search engine space that’s dominated by one player, that’s pretty astounding.
A top 100 website, the engine is ranked #4 in most countries and #3 in Australia. “Some third party estimates say we’re at 50 million users per month, which would be about 40 million searches per day.”
What’s Next for DuckDuckGo?
How does a search engine that doesn’t track customer data make money?
“Search is unique because Google still makes money off search without having to track much ,” Weinberg says. Instead, Google conducts contextual advertising, which displays ads based on what a user is searching (versus behavioral advertising which uses customer data to display ads).
DuckDuckGo can do the same contextual advertising without having to track any data. Calculating customer lifetime value (LTV) isn’t as simple, though. “It’s is even more difficult because of the lack of tracking,” Weinberg says.
Instead of relying on customer data, the company measures factors like brand awareness and market share. Through national surveys, DuckDuckGo asks users if they’re familiar with the brand, how they heard about it, and if they associate it with privacy.
Feedback is an important tool at DuckDuckGo. As they’ve expanded their product line, the team conducted primary research on privacy and people who are interested in privacy.
“We ran different methodologies, national surveys, user tests, and diary studies,” Weinberg says.
DuckDuckGo’s diary studies involved a small group of 12 to 15 people who adopted the product and kept a diary for a period of two weeks. The team would then check during those two weeks to see how the subjects were using, navigating, and feeling about the product.
“We found that would use DuckDuckGo but then click off to other websites that can track,” Weinberg says. “They felt unprotected, and we realized search was only part of the solution.”
This inspired the company’s latest release, DuckDuckGo Privacy Browser, including extensions for desktop that help block trackers and enforce greater encryption across the internet.
“It’s all about encryption and education,” Weinberg says. “We’re trying to simplify privacy.”
Interview by Nathan Chan, feature article reprinted from Foundr Magazine, by Allie Decker
- A history of Weinberg’s early startups and side projects
- The birth of DuckDuckGo and its mission to make the internet less creepy
- Why Weinberg decided to write a book on mental models, and how he uses them on his own team to build better leaders
- Why DuckDuckGo eschews traditional management hierarchy
- The marketing strategies Weinberg used to help DuckDuckGo achieve viral, organic growth
- An overview of DuckDuckGo’s monetization strategies
- How DuckDuckGo used diary studies as a methodology to gather insights
- What’s on the horizon for DuckDuckGo
- Weinberg’s advice for entrepreneurs who are competing against behemoths in their market
Full Transcript of Podcast with Gabriel Weinberg
Nathan: The first question I ask everyone that comes on is how did you get your job?Gabriel: How did I get my job? Well, I started doing start-ups right out of college, and so I kind of formed my jobs over time. And I actually had run a couple start-ups, one failed, I mean, a tonne of side projects, one was pretty successful, and then I really tried to craft this last startup, DuckDuckGo, to be something I really wanted to kind of work at for a long period of time. And so I thought about it a bit more deeply, about what I was really interested in and that kind of thing, and so I guess I’ve been crafting this job for the last 10 years. Although, I must say, I was completely wrong about what it entails, and it changes every 18 months.
Nathan: There you go. So before you started DuckDuckGo, can you tell us about some of your early businesses?
Gabriel: Yeah. So right out of school I started an educational software company, it was kind of a social endeavour, the idea is to raise student achievement through parental involvement, by helping connect parents to teachers through the internet. It was quite literally 20 years too early, or 15 years, depending how you count it, and so it didn’t go very far. This was in 2000, 2001. And then after that I started kind of an early, almost connecting people, social network, but it wasn’t really a social network, it was more like finding old friends and classmates, kind of pre-Facebook, and that ended in 2006. And then I started DuckDuckGo a year later.
Nathan: Yeah. Wow. Okay. So DuckDuckGo’s been going, okay, wow, since 2007.
Gabriel: Yeah. Starting working on it in 2007, incorporated beginning 2008, and then launched at the end of 2008.
Nathan: Yeah. Wow. Yeah. Now, for our audience, would you be able to, I guess, give a short 30-second kind of elevator synopsis of what DuckDuckGo is, for the listeners that haven’t heard of it?
Gabriel: Sure. DuckDuckGo is a internet privacy company now, that really helps you protect your information online. We like to say the internet shouldn’t feel so creepy, and protecting your information should be as easy as closing the blinds, and DuckDuckGo provides kind of the tools and technologies to help you do that. We started out with a search engine, and that is the primary tool we offer, which is a non-tracking alternative to Google. But we also offer a privacy browser on iOS and android, and browser extensions that kind of block trackers and do more encryption across the web.
Nathan: Interesting. So you guys are … Yeah. I’ve known about you guys for a long time. I first heard about you from your book Traction. I don’t know if you remember, but you sent us like 10 copies or something. Do you remember that?
Gabriel: No. No. I do not. But you’re welcome.
Nathan: Yeah. A few years back, I can’t remember why, but yeah, you sent us a few copies, and I read the book, some people in my team have read the book, incredible book. And, yeah, incredible book if you really want to know marketing. It’s just a great book, because I think when it comes to marketing, it’s like you’ve just got to … especially when you’re trying to get “traction” for your start-up, or trying to find a marketing channel that works, when you’re just starting, you’ve built this awesome product or service, you’ve just got to throw things against the wall and just see what sticks. And, yeah, I thought Traction was an incredible book to really just give you a tonne of different ideas and things to try. Like one of those things must work, else there must be something wrong with your product or service. Would you say that?
Gabriel: Yeah. No. Thanks for that. Yeah. I mean, it really is taking a scientific mindset, experimentation approach to marketing, and the book really systematically lists all the 19 channels you could use to get traction. And it really offers the advice, like you said, to just not leave anything out. Really, you got to try a lot of things. And often the things that are kind of unusual and not often working in your industry might be the thing that actually works.
Nathan: Yeah. That’s so true. So while we’re on the topic of books, I definitely want to come back to DuckDuckGo, I really want to hear more about the model, hear everything about this incredible company you’ve created, but let’s talk about your latest book, Super Thinking. What is the premise of this book? I haven’t got a copy yet, but I’d love to hear kind of why’d you write this book, why are you super pumped about it?
Gabriel: Yeah. Well, I need to get you a copy then. Yeah. So the book is about mental models, which is concepts that you really should know to help you be a better strategic thinker. And there are tonnes of these concepts. So for example, in start-up world, people analyse business models with critical mass or network effects, like does this business have … if you reach a certain number of people, will it start to take off or reach a tipping point, which is another mental model. And what I realised, in building out DuckDuckGo, is that to be a really good executive, and especially a founder and CEO, you really need to know all these different mental models, but they don’t generally teach them in school. And so I started to train our executives, and have a whole executive team be good at strategic thinking, and I realised that there was a huge gap in knowledge for a lot of people. And so I literally wrote down a huge list of these things and said, “How many of these do you know?”, and most of the people didn’t know more than half of them.
And that was really the genesis of the book. First it started as this list, and I was like, “Well, go learn them. Here are all the Wikipedia links to get you started.” But then I realised that was kind of inefficient, and a lot of them are interrelated, and so I started working on this book with my wife, actually, who is a PhD statistician. And we systematically kind of list and explain about 300 mental models, through nine chapters kind of grouped together in themes, and it’s why it’s called the Big Book of Mental Models.
Nathan: Yeah. Wow. Interesting. So I’d love to delve more into that. So there’s 300 different mental models.
Gabriel: Yeah. So you can think of them grouped in different chapters, and the chapter-grouping themes would be kind of familiar to you. So the first chapter’s all about biases, like the different cognitive biases that trap you up while you’re thinking or kind of relying on your intuition. So that’s like confirmation bias, but then things that you probably haven’t heard of, like Hanlon’s razor and things like that, which, by the way, is the idea that oftentimes you misinterpret people’s behaviour and you think that they may be doing something negative, when in fact they’re just taking the path of least resistance. And Hanlon’s razor is, “Never attribute to malice which can be explained by carelessness,” which is a very good model for a lot of work interactions.
So then the second chapter’s kind of on unintended consequences, and the third is on how to spend your time wisely. And then so it kind of goes through in these different themes throughout the book. And the last two chapters are very more business-themed. Chapter eight is on management, and how to lead people, mental models. And then chapter nine is on business models and kind of how to get your business off the ground, essentially.
Nathan: Interesting. So you talk about strategic thinking, how can you apply that to a mental model?
Gabriel: Yeah. I mean, so the idea is, what we kind of say in the introduction, is if you think back to when you first learned arithmetic, right, so addition, subtraction, multiplication, division, you start out with addition, and then you learn multiplication, which is basically based on addition, it’s just repeated addition, but if you didn’t have multiplication and you wanted to kind of multiply a big number with a big number, it takes a long time if you just had addition. Say you want to do 543 times 342, I’ve no idea what the answer is, but if you had a calculator in front of you and could use multiplication, or even write out the multiplication, you could do it very fast. If you needed to do addition repeatedly, all those times, it would be very slow. And that’s what mental models really is. Once you kind of know something like critical mass, you can think in a higher order way really quickly, whereas if you didn’t know any of these mental models, you’re kind of starting from scratch every time, and it’s very difficult and time-consuming to make good decisions.
Nathan: So a lot of these mental models, do they help you become a better leader, to understand people and why they do what they do?
Gabriel: Yeah. Exactly. So the first chapters are all about biases, but the eighth chapter is really all about leading people, and so there’s a couple of really big themes in there. One is people are really different, and if you want to manage people effectively you have to understand that each person has a different array of characteristics, and strengths, and weaknesses, and so there’s a bunch of mental models to help you with that. We’re familiar with personality types, IQ versus EQ, kind of your emotional quotient and emotional intelligence, and then all this kind of comes together in this mental model of 10x teams.
And so what we say in the book is … There is another mental model called Joy’s law, which is all the smartest people work for somebody else, which means you can’t really hope to get the smartest people in your company, just because the world is big and there’s so many companies competing with you. And so what you really have to do, is you have to arrange … That sounds like a bad thing, but in fact, what we’ve found out is if you can arrange people in just the right way, and give them the right jobs that fit their kind of characteristics and personality, they can reach extraordinary success, both individually and as a team. And we call these kind of 10x teams, off the myth of the 10x engineer, if you will.
And the mental model to apply in this area is called resonant frequency. And so it comes from physics, and the idea is … You’ve probably seen a movie or some TV show where a singer breaks the wine glass if they sing at just the right note. What’s happening there is they’re hitting just the right energy frequency, where the vibrations are being absorbed by the glass and then it breaks. Well, the same thing can happen with people and teams, except not breaking, it’s just they will absorb the energy that you’re giving them from the right role and job description where they can really succeed. And if you can craft those roles and teams in just the right way, then you can get these teams that just reach resonant frequency and just do amazing things with very few people.
Nathan: Yeah. Fascinating. So it begs me to ask the question, how would you know if the right person is in the right role unless you test, and how would you know that the team is a 10x team unless you test?
Gabriel: You’re exactly right. So the way you know is from output. And so generally, you’re not having people at this resonant frequency at all times, just because it’s not possible to stay in forever, but you really know when you see it, because the productivity of the team and the person is just off the charts, right, they’re getting things done that you would never expect one person or one team to get done. But you’re absolutely right, the way to kind of get and craft those teams is you have to move people around and you have to test.
And at DuckDuckGo we’ve really taken these things to heart. So we don’t have a traditional management hierarchy. We’ve kind of split all our leadership roles into different positions, and we run the company by objectives, and people move around in those objectives, depending on what we’re trying to get done. And we’re constantly moving people around to kind of fit what they’re most interested in and best suited for, in hope to achieve these kind of 10x teams. And we’re only 65 people competing in the kind of biggest market in the internet, if you will, the search market, and doing decently well, and I think it’s largely because of this structure.
Nathan: Yeah. Wow. Fascinating. So was this an idea that you came up with, or did you learn … So let’s just say you do not have a VP of marketing, or a CMO, like that doesn’t exist, right? Or it might exist, but that person could change?
Gabriel: Yeah. That’s right. In fact, we don’t currently have that position. And we’re not opposed to having certain positions, but even if we had that position, they might not be … not all marketing people may directly report to them, they’re not necessarily managing everything. They would be like more of a spiritual leader of that piece, but each objective, each project, each anything has its own owner that may not be them at any given time. But you’re right on that particular one, we don’t actually have a CMO.
Nathan: So how did you learn … Did you come up with this kind of management, I guess-
Gabriel: The short answer is not really. I mean, so all of these things are out there. A few we came up with, but really, the vast majority, I want to say like 298 out of 300, are known concepts, they kind of have Wikipedia pages. This idea of resonant frequency is from physics. And what you’ll find is, and this is kind of why I wanted to lay out the book, is a lot of these mental models come from different disciplines, like scientific disciplines. You might learn that in school as it applies to physics, but some of those concepts from different disciplines, physics, economics, statistics, could really apply well beyond their discipline. And the idea behind mental models is to take a multidisciplinary approach, and take the best ideas from all the different disciplines, and then combine them and use them to do general strategic thinking.
Nathan: Yeah. Fascinating. So I’m just blown away by the way that, I guess, DuckDuckGo works, in the sense of how everybody doesn’t have particular roles in the business, or it could change, and it’s really driven by the projects or the work that is working on, so that could change over time to really try and optimise for the ultimate kind of, what you call, 10x teams. So do you have that? Are you guys achieving that, would you say?
Gabriel: Yeah. I think so. I mean, again, you can’t achieve it all the time, because things change, and especially in our business things are always changing, and so you have to constantly kind of rearrange things. And also people can get burned out, right, that’s another mental model. But, yeah, I think we do achieve it in different times and different places. And our goal is to achieve it, right? So I think if you have a goal to achieve that, then you approach management in different ways, and what you’re really doing, the role of management is to help people succeed, right, is to find them the right roles to really help them succeed.
A part of that is ownership, and so there’s another mental model in chapter eight, that we actually borrowed from Apple, called directly responsible individual, or DRI for short. And what that means is that person, you’re assigning a task, or a project, or a whole objective for the company ownership to that person, and so they’re directly responsible, they’re the directly responsible individual. And what that avoids is something called the bystander effect. And this is very prevalent in a reply all email, if you will, that’s sent to like five people, and you ask a question, and no one replies, because everybody’s kind of a bystander, no one knows if they’re responsible so no one replies. The same thing happens at the end of meetings, when you have action items. But the same thing also happens for ownership of projects, objectives, and everything else.
And so what we try to do in the company is really empower one person for every piece of the company, to say, “Okay. You’re the owner of this thing, and so you need to make good decisions for it.” Now, it’s not that they’re the only decision-maker ever, like if they kind of make a decision and people disagree, people can step up and say, “Hey, I think you got this wrong here,” but they’re the real owner of the day-to-day decision-making, and really taking that thing forward. And once they feel ownership, then it unlocks all sorts of opportunity, no longer are people kind of waiting for other people to give their approval, they don’t need to go up this management chain. And so once we had that in place, we realised we don’t really need the management chain, right? As long as you are assigning ownership to everything, then it doesn’t really need a hierarchy, of sense, and that was kind of aha moment in this whole thing.
Nathan: Yeah. I really like that. There’s this concept, I think it’s by Jocko, where he wrote a book called Extreme Ownership, and I haven’t read the book, it’s on my to-read, and now Super Thinking as well, but I was told about this concept around extreme ownership, and I think that it’s a really important thing in organisations, especially when they’re scaling, because you just got so many things going on, you’re trying to move fast, you’re hiring all these people, you got different projects going everywhere. So this concept of extreme ownership, and I think, yeah, really instilling that in the culture and the heartbeat of, I guess, your team’s operating system, I think, yeah, that’s where it’s at, because I think a lot of the times things aren’t moving as fast as you would like, or they’re just not happening, it’s because, yeah, there’s not that ownership piece.
Gabriel: Yeah. That’s right. One thing that we’ve done to operationalize that is, so in addition to having objectives and projects, and all those have an owner, and people really like that book internally, by the way, we have a whole list of every area of the company, that we call areas of responsibility. And so we run a search engine, so that would include everything like the images, and the maps, and the news function, and the spelling, and all the different things you can think of for that, but also every other area of the company, like in HR, running payroll, and all those different things.
Most of these things are not day-to-day jobs, because they’re on maintenance mode, like we don’t have an active project that is currently being worked on there. Nevertheless, we assign an owner for every single area, because things do come up. And so this helps in two ways. One, when things do come up, that person’s in charge of kind of triaging and deciding whether this is critical or not, whether I need to make a project proposal or not. The other is that for everyone else in the company, you know who to go to for any particular thing. So you go through this large list, you search it, and you can see, “Okay. X and X name is by there, so I know exactly who to go to.” And it’s a good way to operationalize all of this kind of ownership.
Nathan: Yeah. Wow. That’s really interesting. I’m writing that down. So basically, you guys have a directory for every component of your business, each person has ownership. And could one person have more than one thing?
Gabriel: Yeah. Absolutely. In fact, many do. And we kind of organise in objective teams and functional teams. Objective teams are where the work gets done, and functional teams are front end, back end engineering, operations, design. And they more get together for kind of best practises and talking about their craft, but one of the things they do is organise their AORs. So design team has maybe 20 different AORs where they split up all the different design aspects that need maintenance, “Who’s going be the point person for making graphs on the blog?”, and, “Who’s going to be the point person for making presentations, or ad copy, or ad creative, when we need it?”, and so all that’s kind of written down.
Nathan: Yeah. Fascinating. Okay. Awesome. Well, look, let’s kind of switch gears. It sounds like an incredible book, especially it sounds like a really different kind of book, it’s not the run of the mill kind of management, or leadership, or strategy-based book, so no, I’m looking forward to checking it out, I’d love to get a copy. So let’s kind of switch gears. Would love to hear more about DuckDuckGo, how you’ve grown it. Would you be able to share, kind of with the audience, traction thus far, and just put some numbers around it, because it’s a decent search engine in terms of volume of searches every day and users, right?
Gabriel: Yeah. Absolutely. So, man, it’s grown a lot in different orders of magnitude. Right now, and our traffic is public at duckduckgo.com/traffic, we do a bit over a billion searches a month. We’re the number four search engine in most countries in the world. Actually, number three in Australia, for what it’s worth.
Nathan: There you go.
Gabriel: Yeah. And so we don’t track our users, so we actually don’t know how many kind of unique people are using it every month, but some third party estimates put us at about 50 million visitors a month, and yeah, that works out to about getting close to 40 million searches a day, for what it’s worth.
Nathan: Yeah. Wow. And in terms of Alexa, you guys are in the top couple of 100, I saw, maybe top 100.
Gabriel: Yeah. I think we’re getting close, or have broken, somewhere around the top 100 sites on the internet. In most markets we’re in the top 100, for markets that we’re doing well, and probably Australia.
Nathan: Yeah. Awesome. That’s impressive. So talk to me, I guess, around how have you grown this thing, especially when you’re competing where there’s a lot of other search engines, and there’s one that’s just totally ingrained in us, the big behemoth Google, it’s got its own acronym, right, like, “Did you google this?” Yeah. So how do you compete with something like that?
Gabriel: Yeah. I mean, so it’s funny, because we’ve gotten so much bigger, but we’re still tiny in the search engine market, right, because it’s such a big market, because literally everyone on the internet uses search. And so running different stages of growth, just goes back to the Traction book, required kind of different channels. So when we were earlier, we used different things, we used kind of social ads, we did content marketing for a while, we focused on PR for a while. All those things kind of ultimately reached…returns. And for the last few years we have been carried really effectively by organic kind of viral growth, that we tried to push along and kind of make better, but it was basically that, which is kind of offline friends and family sharing.
But now, in the last nine months, we’ve been able to … We were constantly running experiments over that time, trying to figure out how to grow faster, right, because we’ve been profitable since 2014, and we’ve been trying to find ways to use our profits to grow. In the last nine months or so, we’ve been successful, actually, at doing more general brand marketing that has raised our market share. And interestingly, the form that that’s taken, is I wrote some really long-form articles, originally posted on Quora Answers, about why you should use DuckDuckGo instead of Google, kind of how tracking works on the internet, and its harms that it creates, and how you can avoid it, and these articles got such high engagement, more than anything else we’ve ever done. It’s counterintuitive, because you think, “People on the internet, oh, they don’t want to read long-form things,” but it turns out, I think they do if the content is compelling and right. And so now we’ve been finding ways to advertise those answers, and that tactic has worked well in terms of growth.
Nathan: Yeah. It’s incredible, because if you think of in sheer volume of users, not that you guys track them, but if you’re generating 50 million searches, you’ll probably get, like you said … sorry, 50 million monthly uniques estimate, yeah, you wouldn’t think that blog posts … obviously, you guys, your reach is compounding, but you wouldn’t think at your stage you would be doing blog posts and stuff like that, long-form, you wouldn’t think you, as the CEO, going and answering messages on Quora, right, or queries on Quora. It’s crazy. Yeah.
Gabriel: Yeah. You have to think of the scale, though, so yeah, I agree, it’s an interesting tactic. But what it came down to is we’ve only made seven answers and we’ve been promoting them to hundreds of millions of people, via advertising channels. And one of the ways we’ve been doing that is we were the first native advertiser for Quora itself, and so the posts have been promoted on Quora to a lot of their audience, which is about 150 million uniques a month. And another one is Reddit, which we’ve used a lot, which also has hundreds of millions of uniques a month. And so it’s really finding kind of bigger audiences, and putting content in front of them that is actually really compelling, that’s kind of more native to that platform, and that combination has been working. But like you said, the scale is really high, so it’s not cheap, right, at this point, but it is working.
Nathan: So it sounds like you’re not doing direct response stuff, though, like it’s more brand-
Gabriel: That’s right.
Nathan: Yeah. So if it’s more brand, how do you guys work out your LTV, or how do you even work that out per … like your ARPU?
Gabriel: It’s even more difficult for us, because it’s hard to track these things, because we can only do anonymous statistics. But what’s been interesting is twofold. One, is we started out doing this for actual brand awareness, and so one of the ways we were measuring that was running national surveys, and so just how many people have heard of DuckDuckGo, and, “If you have heard of it, can you recall that it’s associated with privacy?”, those are kind of the basic questions. And then we also ask, “If you have heard of it, where did you hear about it? Have you seen it in any of our ads on these various platforms?” So that was one way that decently measured brand awareness in general.
But what we found out is the countries that we were advertising in, our market share was going up more quickly and certain traffic growth was going up more quickly than other countries where we were not. And so partly we’ve been doing this on a country by country experiment basis. And I would say it’s not the tightest loop, in terms of can we get it down to the cent, no. But can we ascribe that this marketing campaign is having this general effect? Yes. I think we can say that definitively now.
Nathan: Yeah. Yeah. Interesting. So one thing I’ve always known DuckDuckGo for, and I think this is your strongest differentiator, and just with basically everything that’s happening with society and the world, and what’s happening in the media, around everything privacy, that’s your guys’ biggest differentiator by far. It is a very, very strong differentiator, with kind of the way that things are moving with the internet and privacy. So inherently, it’s a very attractive alternative, because people don’t like the power that the Googles and the Facebooks have, so you guys have a very nice alternative, which I think is a very, very strong differentiator.
Gabriel: Yeah. I mean, that’s definitely true. We started out with some other differentiators, but that one really is the one that stands out, right? And partly because, like you said, it resonates extremely well with people, especially in the last two years, but also because it’s hard to copy, right, it can actually let us stand out, whereas kind of Google’s business model is built on that kind of tracking that people are looking to escape. And so the percentage of people, which is growing, who really want to escape that tracking, they really have to look elsewhere to go get that, and so we offer that alternative.
Nathan: Yeah. It’s really interesting, and I’d love to talk about your model, because the way that Google builds its inventory, because it is tracking people and their behaviours, makes it very, very strong for a business … to be honest, like a Foundr, like our company, Foundr, we go pretty hardcore on direct response advertising, and using Facebook’s Lookalike Audiences, and all sorts of things, and that AI stuff, and it is so powerful, and we get really strong returns, really positive returns on the front, no earn back periods. So, yeah, that’s kind of the catch 22 in a sense, because yeah, you guys can’t use that, I guess, model to monetize, because you can’t track what people are doing. So I’d love to hear about how you guys have worked out the monetization side of the business. And you guys are profitable, you have been for a few years, which is impressive.
Gabriel: Yeah. What’s really interesting about search in particular, is it’s somewhat unique in the sense that Google still makes most of their money off of Google Search, without really needing to track you much, because the search engine ads are built off of key words. So if you type in car, you get a car ad, and you can do that, because it’s contextual advertising just based on the contents of the page, not based on you as a person, which we would call behavioural advertising, the kind that kind of follows you around, that people sometimes find creepy. And so we can do the same contextual advertising really without knowing anything about you, without tracking you, and that’s essentially how we make all of our money.
Nathan: Oh, yep. So, yep, I’ve got you. Okay. So you guys are really starting to build out your … you have an advertising platform?
Gabriel: So we are part of … There’s really two big search ad platforms. There’s Google’s of course, and then there’s the Bing, Yahoo, kind of DuckDuckGo one, and so we’re part of that second one. And I don’t know if you’re using it, but it’s relatively robust as well, hundreds of thousands of advertisers are on it, and you can get decent volumes on DuckDuckGo, Bing, Yahoo, kind of all the rest of the search engines, basically.
Nathan: Oh, I see. Yeah. Okay. Interesting. So talk to me around kind of … Because what you’re doing, Gabriel, is really fascinating in the sense that you’re taking on this big behemoth company, and you guys are doing it in a really smart way, like the differentiation, and what you’re saying is right on the monetization front, so what’s kind of next?
Gabriel: Yeah. What’s interesting is we started out in search, right, and people definitely want an alternative to Google, and are looking to escape, but they’re really looking for something broader on the internet to protect themselves. What we really discovered is people just want to feel safer on the internet, and they’d like it to be simple, if possible, and seamless, where it’s not kind of creating a sacrifice for them. And search is just one component of that.
And so what we did last year, is we essentially updated our browser extensions and mobile apps to do more than search, and so they also block trackers across the internet, they make you go to the most encrypted version of a website, when it exists, and they give you privacy grades, so they tell you how kind of protected you are at a glance. And when you add all this up, it’s really just in one download, all the privacy essentials you need to protect yourself online, and it’s available on any platform. So you don’t really need to … Kind of before this existed, you had to go figure out all this complicated stuff, and download multiple apps and extensions on different devices. And what we’re really trying to do is simplify privacy for people, so that on any platform or device you just download DuckDuckGo and you’re basically protected as much as you can be while still being completely seamless.
And so that’s effectively also what’s next, because we’re improving these products, and we’re hoping to build even more kind of privacy essential features within them as the landscape changes, and it’s kind of ever-changing.
Nathan: Yeah. Wow. That’s really smart. So when it comes to, I guess, funding, I did see that you guys have raised venture funding. How much have you raised thus far?
Gabriel: We’ve only raised about $13 million.
Nathan: Got you. And you guys are based in Pennsylvania?
Gabriel: Sounds like Transylvania or something, right? Yeah. I’m based in Pennsylvania, but we’re actually a distributed company, so we have people all over. So we’re about 65 people, and we’re completely remote, and so we really have people kind of everywhere across the world.
Nathan: Yeah. Wow. Awesome. Yeah. We’re a hybrid. It’s a combination. So I guess when it comes to, I guess, working out your path to growth around the monetization front, how do you plan to monetize these free apps that you’re building?
Gabriel: Yeah. So they have search as its component, and so when you use our browser or our browser extension you’re also using DuckDuckGo search, and that really is the monetization engine that can power the rest of it. And so we’re kind of lucky in that regard, that search is such a good business model without tracking anybody, and we can use that to kind of fuel and fund all the rest of these privacy essentials, and so we can offer this nice package to the consumer, but the monetization can still be free and stay with search.
Nathan: Yeah. Wow. Interesting. And I think that’s a genius idea to build tools and really speak to that market, because it’s a certain kind of person that doesn’t like to be tracked, there’s a certain kind of person that doesn’t like Google, Facebook, they don’t like ads, they hate pop-ups, all that, they’re concerned about their security. How did you guys, from a product development perspective, work out that that was the next move for you guys, because you could’ve done many different things?
Gabriel: Yeah. That’s a great question. We actually did, and started doing for the last few years, a lot of primary research on privacy, and on people who are interested and who want to act on their kind of privacy interests. So we ran a lot of different methodologies to kind of come to this. We kept doing international surveys. We also ran kind of various user tests, and diary studies, and user
Nathan: What was a diary study, sorry?
Gabriel: Diary studies are where you kind of get someone to adopt a product, and you kind of check … The reason why they’re called diary studies is because you ask them to keep kind of a diary of their experiences. But you effectively check in with them a lot during like a two-week period, and you really try to understand how they’re feeling about the product, is it really helping them, what are they really missing, that kind of thing. And so we’ve done those with DuckDuckGo search for a long time, from the beginning, like take people who’ve never heard of DuckDuckGo search, and then ask them to adopt it from Google, and really try to understand kind of what they’re searching, what they’re feeling, how they like the product.
And so we started expanding on that and asking them, “Okay. Is this satisfying your privacy needs? Why or why not? What else would it need to do to do that?” And one of the main things that immediately became obvious is, okay, you search DuckDuckGo, it’s anonymous and no one can tie your searches back to you, but as soon as you click off the search results, now you’re on other websites that could then theoretically track you, and a lot of them have kind of trackers hidden on them. And so people felt, once they kind of realised all that’s going gone, and a lot of people don’t realise trackers until you tell them, they feel a little unprotected, like search is only part of the solution. And I think that’s really true.
And so what we really set out to do initially was kind of round out the solution, so while you’re also browsing, especially off of DuckDuckGo search results, we’re still protecting you as best we can. And then it came, “What else can we add to that, that really rounds out that protection?”, and so we came up with a large list of other kind of things we could build in that package, and then went similarly, and went and asked people, through surveys, and these diary studies, and other things, to figure out which ones are most compelling to them. And the one that came after that was the two that we built, which was going to the kind of encrypted versions of sites where possible, to make sure your ISP can’t take your browsing history, and then also education of telling you kind of what trackers are on the page, or what are we blocking, that kind of thing. But there were lots of others too, and those are kind of the other ones that we might build in the next couple years.
Nathan: Yeah. No. My mind’s running crazy. You could do a couple of different tools, like a VPN software, you could do antivirus, there’s tonnes of different things you could do.
Gabriel: Yeah. That kind of stuff. Password management, data breach notification, all sorts of stuff.
Nathan: Yeah. Yeah. This is cool. So you talked about these diary studies. I’d like to delve a little deeper before we work towards wrapping up. On average, how many people would you be speaking to every month, and getting them and catching up with them for a two-week period, to get this kind of insight, qualitative feedback?
Gabriel: Yeah. You could think of a lot of these methodologies as on a spectrum, right? So on a national survey we might do 2,000 people, right? In the middle, we kind of use other things, like PickFu, which is a quick polling thing, it might be 50 or 100 people. And then diary study’s all away on the other end, and it’s usually we would kind of recruit like 20 people, it could end up being as little as 12 or 15, and it’s really in-depth, right? So you don’t have many people, but you’re going really deep with them. And you want to get enough people where you can start to see patterns, but because it’s so in-depth, you don’t want too many people, because it’s overwhelming, because one person at our company is running the diary study. And so they’re literally having interviews, and so they’re checking in, I think, with everybody, I mean, it varies per study, but generally once a week.
And depending on the study, we’ve created tools to help them make the diary. So at one point we made a browser extension that helped them take screenshots of the searches that they’re having trouble with or have comments about, and so you could just click a button and type in a box, and it would kind of all collect that for us. And when they meet weekly they would go over that and kind of drill down and see what they were thinking at that time. And then the person running the diary study would then be compiling all these things, and trying to make sense of all the data, and be like, “Okay. Here are the themes where people really got tripped up. Here are the different personas. Here are the percentage of people that said they would actually stick with DuckDuckGo after the two-week period, and here’s the percentage that I actually believe of that,” that kind of thing.
Nathan: Yeah. Wow. Interesting. No. This is great. Okay. Well, look, we have to work towards wrapping up, but that was gold on, I think, people can learn around how to really get this feedback and see these patterns to build great products. So two last questions, Gabriel. One, where’s the best place people can find out more about your work, and also your latest book, Super Thinking? And by the time this comes out the book will be live, so where can they go get a copy? And the second one was, I guess, for anyone that is about to embark on building this crazy start-up, and, particularly in your case, taking on the 300, 400 pound gorilla in the marketplace, like you guys are, and still keep pushing forward, what do you have to say to them?
Gabriel: Yeah. So in terms of connecting with me, you can find me probably easiest on Twitter, @Y-E-G-G, yegg. For Super Thinking, it’s at superthinking.com, so that’s easy. In terms of kind of taking on the 800 pound gorilla, I mean, there’s a post I wrote a long time ago, which I think other people have written, several as well, is that when you’re starting a start-up, it’s extremely difficult, but actually, if you start a really ambitious start-up it’s not necessarily any more difficult than starting a non-ambitious start-up. And in fact, there are some reasons why it is actually easier, because kind of the more ambitious you are, the more exciting it is to investors and employees, and easier to get kind of other people on board.
And going back to this kind of 10x team notion, is if you’re going to succeed at doing something kind of ambitious and big, you’re going to need an amazing team around you. And so having that kind of idea and mission that kind of attracts those people actually is essential to attracting that team. And so that’s the core piece of advice I’d probably give, is to just work on crafting the company, and values, and mission, to attract that team, which you will need to then go after and kind of complete your ambition.
Nathan: Love it. Awesome. Well, look, thank you so much for your time, Gabriel. This was an incredible interview, and yeah, I, yeah, can’t thank you enough for your time, and congratulations on all your success.
Gabriel: My pleasure. Thanks for having me.
Key Resources From Our Interview With Gabriel Weinberg
- Give DuckDuckGo a try
- Follow Weinberg on Twitter
- Order a copy of Weinberg’s latest book Super Thinking: The Big Book of Mental Models