Scientists should learn from Jeff Bezos

To create truly transformative impact in the world, a scientist must systematically innovate over decades in areas overlooked by others. How can you model yourself after this ideal? You could study Nobel Prize winners, but you’ll risk getting stuck in the scientific details and miss the general themes. Better to examine a non-scientist and capture the abstract concepts onto which you can map your specific circumstances. Enter Bezos. Scientists should learn from Jeff Bezos.

Few people have transformed our modern world more than Jeff Bezos. As the founder of Amazon, he pioneered online retail, created a truly global marketplace for small entrepreneurs, invented cloud computing, and built a logistics system capable of getting pretty much anything pretty much anywhere in less than two days. In his spare time, he built a rocket company and invested in numerous startups—like Google before anyone had heard of it. Bezos has created truly enormous value for the world, and our system of capitalism has rewarded him accordingly—making him one of the richest people on Earth. But Jeff Bezos is more than that and scientists could learn a lot from studying him. I recommend reading Invent and Wander, co-written by the famous biographer Walter Isaacson and Bezos himself.

Scientists should think long-term like Jeff Bezos

Amazon was built on long-term thinking. In his first annual letter to shareholders, Jeff Bezos famously stated:

“When forced to choose between optimizing the appearance of our GAAP accounting and maximizing the present value of future cash flows, we’ll take the cash flows.”

Masked in the corporate vernacular expected from a professional executive, Bezos boldly proclaimed his principled stance that he wouldn’t play financial games to create (and be rewarded for) paper profits. The more impressive part is that he has stood by and lived out that principle for the last 25 years.

There were many times when, under pressure from Wall Street, he could have prioritized short-term profits. For example, when Amazon created the ability for customers to leave reviews—even negative ones—it measurably hurt sales and Wall Street analysts were outraged. But customers loved it. The feature prevented them from buying something that they would regret, and it incentivized suppliers to improve product quality and meet the expectations of customers. In the long run, this built a stronger bond of trust and loyalty with customers and contributed to Amazon’s incredible growth. However, it was not the easy choice or path of least resistance at the time. 

Set clear goals

To make good long-term decisions, you need a clear, internally consistent goal. When you don’t have such a goal, externally inspired, fleeting aspirations will fill the void. If Bezos lacked a true North Star to guide him, a chorus of disparate opinions would have certainly rained down on him—likely advocating short-term approaches. Luckily, Bezos had a clear North Star: always obsess over customers. (More on that later.) 

As the time horizon increases, uncertainty generally increases, but there are also other predictions that become easier. This is why day trading stocks is extremely risky, but investing in a diversified portfolio over decades is a very safe bet. For Bezos, it was difficult to predict which new products and services will catch on—especially in the new world of the internet and digital commerce. Would people be willing to buy books, clothes, groceries, TVs, and time shared computing resources from the same company over the internet? He could make informed, well-reasoned bets, but they were fundamentally uncertain. (In this case, the answers were a mixed bag of yes, sometimes, increasingly so, no, yes, and hell yeah).

You just can’t know how the world is going to change over months and years, much less decades. However, Bezos realized that he could know with certainty that future customers would definitely want more selection, lower prices, and faster service. 

Ensure your long-term thinking translates into actions

Importantly, Jeff Bezos didn’t just think long-term, he aligned his actions with his long-term analysis. Applying the rationale that:

“If you’re long-term oriented, customer interests and shareholder interests are aligned”

he employed billions of dollars of capital. He bought fleets of vans and airplanes, built massive warehouses around the world, and hired huge numbers of expensive software engineers. These enormous outflows of cash were able to create a truly unique and differentiated experience that delights customers. But the expenses occurred well before any potential payoff, and success was never assured. In fact, Jeff spent billions on risky experiments that failed to pay off. When was the last time you saw someone on a Fire phone? That experiment cost more than 200 million dollars. Nevertheless, Amazon has continued to accumulate value, because Bezos had the fortitude to act on his intellectual understanding that one success would pay for a thousand failures. 

As a scientist, how can you improve your long-term thinking? Of course, you think about publications as a series of figures, produced from a series of experiments, perhaps performed over years. And your education and training may have taken decades, but these attributes are generally true of all scientists. How do you differentiate yourself?

What are your guiding principles?

Bezos used customers as his North Star. Is it an academic professorship, a Nature paper, inventing a life changing cure for some disease? These are certainly aspirational goals, but they are not the best long-term goals. If you achieve these ends, what next? Will your pursuit of these goals drive you to make decisions you’ll ultimately be unhappy about (like sacrificing too much time with your family or trading your passion project for one that is likelier to get funded). Scientists should frame their goals in a way that carefully considers their locus of control. You don’t completely control reviewers, hiring managers, the popularity of your field, the productivity of your students, and so on. 

While I can offer some suggestions about the types of goals that might be best for scientists, this is going to be a very personal exercise where you had better be honest with yourself. The stakes are high. Science and scientific training takes a really long time, reducing your ability to pivot or iterate. The cost of chasing the wrong goal compounds very quickly.

What can you be certain of and what can you not know?

Perhaps you are absolutely certain that drugs of the future will be personalized, like the CAR-T cell therapies that have gained traction over the last few years. But as an undergraduate student today would you stake your career on CAR-T cell therapies being the gold standard in 30 years? I wouldn’t. That doesn’t mean avoid CAR-T cells at all cost. To the contrary, ride that wave while it lasts. Just appreciate that it might be a wave that will wash out and leave you behind.

Instead, become an expert in CAR-T cell biology, while gradually gaining expertise in programmed NK cell therapy, PD-1 checkpoint inhibitors, mRNA vaccines, computational biology, and protein engineering. Like a well-diversified venture capital fund, some of these will prove to be transient fads but some will definitely hit it big in the long run. While that’s all sorting itself out, you’ll be aggregating diverse knowledge bases in one individual and become a one-of-a-kind unicorn. A one-woman interdisciplinary idea incubator, you’ll be more likely to find valuable tangent angles to differentiate you enough from the herd of consensus scientists in your field. You’ll be rationally balancing risks and rewards like a disciplined practitioner of portfolio theory in finance–or the part-time venture capitalist and former investment banker Jeff Bezos himself. 

Understand your appetite for risk

If you read that and thought “To hell with hedging. I’m smarter than the herd, and I’ll prove it. Watch me go redefine science and win the Nobel Prize.”, make sure you have a large appetite for risk. Let’s say you put all your eggs in one basket and set out to prove your theory of everything, but the unifying model never materializes. You might spend 10 years in graduate school and never get your Ph.D. You may limp along, getting postdoc after postdoc after non-tenured position but never secure the level of financial security or emotional stability required to start a family. I don’t mean that to cast judgment on other scientists’ choices. To the contrary, I can admire the romantic sense of struggling for Truth and appreciate the entrepreneurial drive. However, own the risks and tradeoffs associated with your choices.

Do the self-analysis early and often, because the opportunity cost of spending years pursuing an academic dream that never materializes can quickly spiral. If you’re still truly convinced the scientific community will meet you in the future you gloriously designed and you’re okay with the associated personal costs, I commend you. Like Elon Musk, you may be rewarded handsomely for assuming the risk and creating something legitimately world-changing. Perhaps it’s true that “the only way to successfully predict the future is to create it.”

Are your actions aligned with your long-term strategies?

As a scientist, you’re likely inclined towards and capable of thinking many steps ahead and creating a detailed long-term plan, but can you stick with it, making it reality day-after-day for years? We’re all human, and human nature can make it difficult to work out regularly, save your money, moderate your portions, or floss twice a day. We all have areas of life where our actions do not match our rational plans and best interests. You don’t have to be perfect.

Be kind to yourself, and accept that you’ll always be a work in progress. However, recognize where your actions are falling short of your plan, and work methodically to close those gaps. Let’s say you hate scoring the hundreds of immunofluorescence micrographs produced during your experiments, so you tend to avoid the analysis which ultimately delays publication of your papers. Do NOT allow yourself to do anything else until those micrographs are scored. You can set up little rewards to incentivize yourself, if that’s your thing. But it’s always going to come down to self control.

You alone are responsible for getting things done. If you can’t bring yourself to do the required work, you have no choice but to admit that you don’t actually really want that long-term goal. Use this logic in both directions to establish internal cognitive alignment. If you really don’t want to do something, you can let that be a signal that perhaps you need to refine your long-term goal. Alternatively, if you’re very certain about that long-term goal, then you have no choice but to do the the task even though you don’t enjoy it.

Scientists should make deliberate decisions like Jeff Bezos

When you’re making big decisions that will affect your career for years to come, you’d better get them right. As a scientist, you make a lot of these decisions. Where to go to graduate school, what to choose for a thesis project, and what to propose in your first R01 grant application are enormous decisions. Get them right, and you’ll stay on the narrow path to your ultimate dream career. Get them wrong, and it could take years to recover if you’re lucky. These types of decisions require a very deliberate, disciplined process.

One-way doors vs. two-way doors

Jeff Bezos’ primary framework for deliberate decision making is differentiation between one-way doors and two-way doors. A two-way door is one you can walk through, then retreat back through the door and be in the same place you started. These decisions are low stakes and don’t need very careful consideration. Including an extra condition in an already planned experiment is an example of a two-way door decision. You’re already doing the experiment, and there’s very little marginal cost in time or money. If it doesn’t work out, you don’t lose anything.

In contrast, a one-way door decision has high stakes. If you take a regulatory affairs job at a big pharma company, you’re probably not going to be able to go backwards and become a professor or a venture capital associate. It’s not impossible but it’s going to be really, really hard. Even though you didn’t spend any money or material resources, there was a big opportunity cost. By pursuing the regulatory affairs position, you likely sacrificed the opportunity to develop the skills and relationships required to pursue those other paths.

Before you make a one-way door decision, apply a comprehensive, deliberate decision making process. While you need to come up with a framework that makes sense for you personally, here is some general advice:

Be patient

Really important career decisions rarely need to be made in a hurry. Don’t let others impose unnecessary time pressure on you. Take your time. Better yet, reflect and strategize regularly, so most of the hard work is already done when presented with an important decision. 

Solicit diverse opinions

Diverse perspectives always result in better decision making. However, don’t make decisions by committee. When it comes to your career, you decide. Nevertheless, trusted advisers can reveal different angles or identify unconscious assumptions. Your best friends may not always be the best advisors in these situations due to their proximity and unquestioning loyalty to you.

Consider multiple time horizons

Don’t make a decision only based on the now. A new job with a raise is always great, but does the decision maximize your career earnings? Does it close doors or restrict your ceiling? Alternatively, don’t completely forego getting paid now in favor of an amorphous promise of a prosperous future.  

Frame broadly and narrowly

Consider your decision from a utilitarian perspective. Is it creating the greatest benefit for the greatest number of people? Putting your entire lab’s names on every publication gets everyone a lot of publications; however, it dilutes the value of an authorship, hurting the highest performers who would otherwise differentiate themselves with more papers. Alternatively, consider the decision very narrowly. Does it really benefit you to stick around and help your thesis advisor finish a grant proposal when you’re ready to move on?

A scientific coach can be a great resource to help you get the really important, one-way door decisions correct. Whether you use a coach or not, give these decisions the attention they deserve. They’re among the highest value activities you perform.

Focus on the important decisions

Bezos understands that getting one-way door decisions right is the job of an executive. He remarked that:

“You don’t get paid to make thousands of decisions every day. You get paid to make a small number of high quality decisions….You need to be thinking two or three years in advance, and if you are, then why do I need to make a hundred decisions today?”

As a scientist, this is definitely true. Think about really prolific senior researchers in academia or industry. Almost by definition, they have substantial teams that actually do the work. As you progress in your career, it’s inevitable that you too will need to focus on a different set of competencies: you’ll need to recruit, motivate, guide, and–above all–make good decisions. You only have two hands and don’t scale, but your decision making can scale. It’s a cruel fact of life that what got you to where you are, often won’t get you to where you want to go. So focus a little less on your technical expertise and a little more on your decision making. That’s a skill that will remain relevant no matter where your career goes.

Maximize the expected value of your experimental portfolio

Experiments are the lifeblood of science and of innovative corporate enterprises like Amazon, and quality decision making can increase their expected value. Experiments are intrinsically risky. They require an outlay of resources (time, labor, reagents, equipment) and also have an associated opportunity cost (performing one experiment means you can’t take on another). Yet, experiments have a wide range of expected values, ranging from negative to very high. While you can never control the outcome of an experiment (you’re a scientist, not a wizard), you can shift the outcome probability distribution.

Deeply considering the experimental design (see the list above of what’s in your locus of control for some considerations for your analysis) can increase the ceiling, increase the floor, and decrease the range of outcomes. This is where an experienced scientific leader can really scale their expertise and wisdom. Take risks while maximizing the expected value of your experiments. That’s how Jeff Bezos built Amazon. He made really big, well-informed bets on experiments. Some failed spectacularly–like the Fire phone. But there were a lot of winners–some world changing–and, on average, Amazon generated a higher return on investment from their experiments than any of their competitors. The immense value Bezos created for himself and Amazon’s investors is a testament to the power of great decision making.

Be active, not passive

Being deliberate means not stumbling passively along your current trajectory. You should aspire to do the most important thing possible everyday. Of course some days you’re just going to be growing cells or preparing solutions. But you should only be doing that, because you’re laser focused on the important experiment it enables.

Every day is a choice and you’re in full control of your choices. Every year, Jeff Bezos empowered Amazon employees by offering to pay them if they quit. He forced them to abandon any lethargic momentum and actively opt in to their job. You too should feel empowered. Don’t fall prey to the sunk cost fallacy—you can quit a project, a degree, a job, or a city on any given day. Of course, there’s something to say about soldiering through adversity and accomplishing something really hard. But a lot of scientists should quit more often. And if you choose to persevere through a really challenging situation, do so actively, fully owning both the consequences and ultimate rewards.

Like Jeff Bezos, you should know who you’re serving and what your mission is

Scientists tend to under-appreciate the fact that they have customers. If you’re a grad student, your customers might be your thesis advisor, your thesis committee, and peer reviewers. For professors, your customers might be your department and your peer reviewers—both for publications and for grants. If you’re in academia, your customers might be your boss and your boss’ boss. All of these are proximate customers. More distant customers may include sick patients or electric car battery manufacturers. If you’re solely in it for the generation of new knowledge, perhaps your customer is humanity or future generations. All of these are plausible customers. Knowing exactly who your customers are can be clarifying and help guide your actions as you try to solve their specific problems.

Be customer obsessed

With Amazon, Bezos was customer obsessed. As the CEO of “The Everything Store”, he knew that his customers wanted whatever they were thinking about at that moment, and they wanted those products cheaper, faster, and of higher quality. This very practical customer definition led to One-Click Shopping and the allowance of reviews (even negative ones). While creating AWS (Amazon Web Services), he had an aspirational image of a customer that didn’t yet exist. Nevertheless, he pictured “a college kid in his dorm room having the same access, the same scalability and same infrastructure costs as the largest business in the world.” This hypothetical customer drove the design of a business unit that realized its ambitious goal and enabled countless startups to blossom. 

As an investor, Jeff Bezos has supported businesses with a clear mission. He tries to discern the mercenaries (those who only want to get rich) from the missionaries (those who are pursuing a purpose greater than themselves). He was one of the first investors in Google who was ambitiously trying “to organize the world’s information and make it universally accessible and useful.” Missionary. He bought the Washington Post. Part of its mission statement is “In the pursuit of truth, the newspaper shall be prepared to make sacrifices of its material fortunes.” Missionary. Bezos bought Whole Foods. Their mission is “to nourish people and the planet.” Missionary. He founded the rocket company Blue Origin whose mission statement includes “to inspire the next generation to enable millions of people to live and work in space for the benefit of Earth.” Missionary. Jeff Bezos has put his substantial means behind important missions.

What’s your mission?

As a scientist, is your mission clearly defined? What are you trying to solve and for whom are you trying to solve it? It’s very easy to get lost down the whole of the problem, experiment, or paper you’re currently working on, but those are not worthy missions. Your mission should be grand, ambitious, personal, and impactful. It should be worthy of a life’s work—not merely the actual labor expended. You should be excited to tell anyone that will listen. Of course, you’ll have days where the work doesn’t feel so motivating, but the mission should be important enough to motivate you through those hard days.

In 1997, Bezos wrote

“This is Day 1 for the Internet, and, if we execute well, for Amazon.”

He meant that it was just the start, and they were excitedly and actively attacking every day. He continued to foster this zeal for the next 25 years, repeatedly saying that it was still Day One at Amazon even as the years passed. This is what he wanted. In Bezos’ mind,

“Day 2 is stasis. Followed by irrelevance. Followed by excruciating, painful decline. Followed by death. And that is why it is always Day 1.”

As a scientist, you should try to keep the magic alive. Have a mission worth pursuing and relentlessly chase it. Be excited. Stay ambitious. Don’t lose your sense of wonder.  When you’re excited and motivated, every day can be Day One for you too.