Vogue: Moon Walker

Obi Felten leitet die visionärsten Teams bei Google. Im Geheimlabor X entstehen Produkte und Businesskonzepte der Zukunft. Kontrolliertes Scheitern inbegriffen

Obi Felten leads the most visionary teams at Google. The secret laboratory X is where the products and business concepts of the future are created. Well-managed failures included

By Esma Annemon Dil
Photos by Rainer Hosch

Styled by Deborah Ferguson
Cover article of Vogue Germany Business Edition
September 2018

English translation: Moon walker


The secret laboratory X was first developed eight years within Google – a place where experts could push the boundaries of innovation. Think of Q’s lab in the James Bond films – and indeed, internally it is called the “Moonshot Factory”; not because they’re preparing to blast off to the moon, but because, with every new project, the team aspires to bring fantastic ideas to life that are at least ten times better than anything that’s already been done and therefore sound like pure science fiction. Increasingly, this now incorporates visionary entrepreneurial strategies and future-looking business plans, which is why co-founder Astro Teller installed German strategist Obi Felten in 2012 as “Head of getting moonshots ready for contact with the Real world”. Felten’s task is to lead her teams so they figure out as quickly as possible which ideas are worth pursuing and which should be dropped, such as the space lift vision.

Successful projects are then spun off as separate companies: Waymo, for example, or Calico – a research lab exploring health and prolonging life, Google’s Brain AI headquarters, the cyber-crime company Chronicle, or Dandelion which looks at revolutionary heating and air-con technology. The Wing and Loon delivery drones, which are balloons that bring the Internet to remote areas, were going through the well-moderated process of separation during our visit.

It doesn’t matter what the sector is – the question will always be whether it’s worth persisting despite all the hurdles, or whether you are using unproductive energy and need to stop. You must have learned a lot about this in recent years.

Yes, but we do reward well-founded failure because it is often the team itself that recommends stopping a project. This is despite the fact that is painful to say goodbye to an idea that is great in theory and one in which you have already invested your heart and soul. That’s one reason for trying to define what we call ‘kill criteria’ right from the start. These may be technical or financial reasons and point to when we have to give up on a new project.

We start on the hardest things first. It is easiest to explain it by means of an illustration. If you want to test whether a monkey can be put on a pedestal and recite Shakespeare, you could spend a lot of time selecting fine excerpts for it to read and build a really great pedestal, but the bottom line is: can you actually teach a monkey to do something like that at all? When I’m discussing the work plan with my teams and everything we need to get done, the first thing we ask ourselves is: what’s the monkey here? This is so that we don’t get tempted into starting with the pedestal first.

So problems first, then the fun bits later? On my own ventures into the world of entrepreneurship, I dove straight into developing logos and packaging design until months later it became crystal clear that there were far more urgent product challenges that should have been faced.

That’s a classic! You’re not alone in this, it’s totally human nature to begin with the things that we’re most interested in, that are in our comfort zone and that we can most easily tick off as done. This gives us a warm glow of satisfaction, a feeling that we’re in control and making progress. At the same time, most companies and investors are more concerned with the figures, the number of milestones on the list they have reached, rather than whether the critical issues have been resolved. Things like branding or packaging would totally freak out the engineers in my team, they would much rather focus directly on the code and would probably speed off in that direction. That is why it’s so important that a team has a mix: people with different strengths, who see things from different perspectives and have different priorities, who can quickly figure out whether there is a production partner you can use or whether the idea is technically viable.

So many compelling reasons for team diversity, yet it still seems so difficult to achieve.

Absolutely. In many companies, alliances are formed based on how much easier it is to connect with someone you share experiences with, who attended the same college or studied the same subjects, has similar views to yours and who confirms your own prejudices. I go out of my way to avoid hiring people who are too much like me, because I simply won’t learn as much from them.

Have you made mistakes because a team’s ‘group-think’ has ended up distorting your own view of reality?

Some, yes; the Wing project springs to mind. We wanted to use our drone to deliver defibrillators in heart emergency situations and get the life-saving device on the spot within 90 seconds of the emergency call being made. The drone engineers were totally fired up by this specific application until one of our experience and research team actually tested how quickly untrained people could use a defibrillator on a dummy. 40 percent didn’t manage it at all, the others took at least five to six minutes – and that was in a controlled environment, without panic or other distractions. Wing would simply not offer any advantages in populated areas, where first responders are on location within about eight minutes. This data clearly demonstrates the importance of looking at a topic from a number of different angles. Engineers just ask: how can we build it? But there are so many other relevant questions around the user experience: Why and how would anyone need to use this?

Or taking a step backwards: Are we even solving the real issue for the end user? We often benefit from productive conflict within the team and from the many different voices all debating around the table. Sometimes you hit lucky and find specialists who can somehow combine different points of view within themselves. For instance, we have a hardware engineer with us who used to be a concert pianist and a firewoman, is a trained paramedic and works as a 3D print artist in her leisure time. But there is yet another very important factor besides the diversity within the team: we only keep projects secret as long as is absolutely necessary, because we always need quick feedback from the real world.

Mark Zuckerberg coined the phrase “Move fast and break things”, shooting from the hip was not only a problem for Facebook though.

I come from the software world, where you can always continue product improvement post launch. Gmail and Chrome were beta versions for years until they finally began to work well. This is not as easy with hardware. You can’t put a part-finished self-driving automobile onto the market, that’s why we have very stringent safeguards at Waymo. However, we often test our concepts early on using test subjects so that we can get feedback as quickly as possible on whether our product is working and whether we are tackling the real issues for users. Many European companies, in particular German companies, find this a difficult mindset to get their head round, as they have a long tradition of painstakingly perfecting the craftsmanship and technology. It may hurt to publish a project that is only 80 percent of the finished article and receive heaps of criticism about it, but these early lessons really are worth it.

How did that go with Google Glass?

It was obvious that Google Glass was too identifiable as glasses with an integrated computer to be unobtrusively worn outside of the building. So we made the product public and tested it within an explorer program in all areas of life: with fashion models, physicians, factory workers, parents wanting to take photos of their children, people on a night out, and so on. The learning curve was pretty dramatic. The battery didn’t last long enough, the voice-controlled user interface was unwieldy, people didn’t find it convenient enough for private use when viewing their tweets hands-free or taking photos for instance – particularly if they didn’t already wear glasses in the first place. Some wearers were thrown out of bars because rumor had it that the glasses automatically filmed everything, such as things going wrong. There was a massive backlash after the initial hype. Of course, this was a traumatizing experience for our engineers, but it was actually incredibly valuable to get feedback at that time and not years later. As it turned out, the project had not completely failed at all, but happened to appeal to a different target group. It is very successful as a B2B application in many industries, in the medical sector, in the service industry and most of all in manufacturing, for example at General Electric or Boeing where assisted reality can increase efficiency in production technology.

Your work comprises so many complex fields of activity, how did you actually come to get this job?

The approach came about internally. I was Director of Consumer Marketing in Europe, the Middle East and Africa for Google, living in London. That in itself was no easy task either, considering that there were over a hundred Google products in these very differing markets. I had some experience with difficult launches, for example convincing Internet users to switch to Chrome, when back then only eight percent of people actually knew what a browser was. When I first met Astro Teller, who is now CEO at X, he told me all about these incredible projects: about contact lenses that measure diabetics’ glucose levels through their tear fluid, or about Loon using balloons to take the Internet into remote areas. It was mostly me asking the questions: Is that legal, flying balloons over different countries? Has anyone talked to any governments about all this? Will these work hand-in-hand with phone companies or compete against them? Do any of these projects have business plans? What about privacy when the balloons are passing through different networks and so on. It turned out that no-one was dealing with these really important non-technical aspects. It was such a unique opportunity, so I figured it was well worth the move to California for me and my family, even though my husband and I had just bought a house in London and our son had just started school – not really the best time for our family.

That was 2012, and just three years later you founded a department at X called the Foundry.

After I had built up a team of non-technicians and systematized all the processes, it made sense to formalize all this. Foundry is the phase projects enter after they have passed their first Rapid Eval – when we know that an idea is, at least in principle, technically feasible. We used to keep ourselves very isolated, but now we form partnerships with industry much more quickly, for example with Telefónica and ATT at Loon and with Jaguar, Ford and Chrysler at Waymo.

Where do the ideas that end up at X originate?

We do not do basic research ourselves, but the Rapid Eval team does read scientific publications and keeps in close touch with research labs. Many ideas start out at universities: Brian Otis, a professor at the University of Washington, gave up his position to make his idea of contact lenses for diabetic people a reality with us. This project is now part of Verily, a company that emerged from X, which has now formed a partnership with Novartis AG. X passed the baton on, so to speak, from a somewhat risky initial prototype out of an academic research lab to a major investor from the pharmaceutical industry.

This year, the “Wirtschaftswoche” magazine presented you with an award as ‘Future Thinker’, and you have also been a member of the supervisory board of the Springer Nature scientific publishing group since April. Would you say that your training has forged the path for your visionary career?

I have not had a linear career path. I grew up in Berlin, studied philosophy and psychology in Oxford in the early 1990s and then started work at the strategy consulting firm Mitchell Madison Group. More and more companies were working on the Internet at that time, and that caught my interest, so I didn’t go back to university to do an MBA after my consulting experience like so many of my friends, but began as a product manager with the eToys start-up company when it expanded into Europe. I convinced the management that alongside the United Kingdom, Germany was an emerging new market for the online distribution of toys and that I should be the one in charge of the project.

You were very young at the time; where did all that self-confidence come from?

Good question, I was 27 and completely inexperienced, but the management at eToys thought it was worth taking a chance on me. I never gave a thought to my self-confidence back then. I think I owe it to my parents though, as my mother is a teacher and my father a professor. As educators, they instilled a love of learning in me from a very early stage. The second formative phase was Oxford, where you learn to gather material in tutorials and long reading lists, and then deepen and develop these ideas further by writing seminar papers. I learned how to learn quickly!

How do you motivate your children to learn?

I have a 10-year-old son and a five-year-old daughter. Of course, this is on my mind a lot. I only hope they’ll carry on being as curious all their lives as they are now. If they gain the ability to familiarize themselves with new areas then they will retain that ability for the rest of their lives. For some time now, lifelong learning has been a buzzword, and it is becoming all the more pressing that we put this into practice as a society. We need a future education system where we take on education continually and not just for a short time at the beginning of our lives. It is of course also becoming increasingly important for us to be able to rate the sources and quality of information more effectively and not just unquestioningly believe what the first link we are sent to tell us.

You once quoted a study by the World Economic Forum which suggested that 65 percent of the professions that the children of today will go on to pursue in the future have not yet been invented. So, how can you possibly prepare them for these challenges?

If the world suddenly changes around you, it is important to stay flexible and deal with the uncertainty. If the world suddenly changes around you, it is important to stay flexible and deal with the uncertainty. Psychology professor Carol Dweck describes the advantage of the growth mindset as compared to a set mindset. The growth mindset is focused on an attitude of continuous growth that will open the door to further development. Children who expect to get everything at the first attempt balk at more challenging tasks – for example, they are less likely to dare to tackle difficult math problems. On the other hand, children who grow up in an environment where the reward is more for experimentation and hard work rather than perfect results tend to have the courage to tackle complicated tasks and spend longer trying things out. In most cases, these people are better able to deal with disappointments, learn from their mistakes and start over again. My own parents gave me exactly this freedom to maneuver, where it is okay to try things and, at times, get them wrong.

You spoke in public about your failure at eToys and emphasized the value of this experience with the benefit of hindsight.

I built the company up from scratch and hired buyers, programmers, sales and marketing experts, many of whom had worked years in traditional companies. One week before its official launch, the entire European expansion was pulled by the parent company in the USA on the grounds that it was costing too much. We had a warehouse full of toys, and I had to let go all the staff I had just hired, some of whom had given up their good company pensions to work for me. At least we were able to pay everyone off and, with their experience in the up and coming e-commerce sector, they all ended up finding good positions again, at eBay for example. I have learned that you have to make decisions quickly, and above all that you need to treat people decently and, afterwards, help them land well elsewhere. That doesn’t mean I’m not afraid of losing, though. Fear can protect you from excessive optimism or, worse, arrogance. But ever since then, I have always known that I will not get paralyzed by fear, that I can overcome my fears and deal with mistakes.

What do you see as your greatest achievement so far?

In the past, I might have given a list of companies and products I’ve had a hand in creating or launching, such as eToys or Google Chrome. The older I get, though, the people are more important, those I support and whose careers I can positively influence. This includes many women. When I worked in Europe, supporting other women seemed to me to be less of an urgent issue. There are so many role models of successful women in politics, the media and other sectors. But Silicon Valley is dominated by tech companies, where women account for only about 30 percent of employees. This is why it is particularly important for me to be a mentor for young women over here, within our company, and outside. One of the women on my team was initially shy and afraid to contribute in meetings. It was really great to see her developing more self-confidence with the right support, starting her own project, then making the decision herself to shut it down. She led the next project, Dandelion, to its successful completion, spun it out of X, became a start-up CEO and just raised her second round of VC investment.

So, has a success formula of some kind emerged during all these years at X?

During our conversation, I am aware that many of our success principles seem to contradict general protocols and do not necessarily make intuitive sense. For example, jumping straight to the problem, that controlled failure can be a positive force, that pioneering solutions that are ten times better do not have to be ten times as complicated or expensive. Therefore it is sometimes worth reaching for the stars.

Pullout quotes

“I try to avoid hiring people who are too much like myself”

“We always need quick feedback from the real world”

“I asked questions: Are there any business plans? What about privacy?”

“My parents instilled a love of learning in me from a very early stage”

“Many principles of our success contradict general protocols.”


MASTERMIND Non-technician among nerds. Obi Felten studied psychology and philosophy, but she asks the right questions. Dress: Dries Van Noten. Ring on the left hand: APM Monaco.

PASSION FOR EXPERIMENTS Obi Felten decides whether a space lift or WLAN balloons for Africa will become reality. Dress: Zero + Maria Cornejo. On the right: Dress by C/Meo Collective.

PLAYFUL Google X CEO Astro Teller took the VOGUE business premise “Women in the foreground” literally and photo-bombed himself cheerfully into the background during the photo shoot.

Top and skirt: Dries Van Noten. Big ring on the left hand: APM Monaco. Heels: Pierre Hardy.

Photos © Rainer Hosch. Please visit www.rainerhosch.com to see more of his work.