On Meta-Experiments
Call me crazy, but I don’t believe a company’s leadership team’s most important job is to deliver a positive financial outcome. Yes, I know that technically any corporation’s leadership and board of directors has a fiduciary responsibility to make decisions in the best interest of shareholders, and that often gets interpreted as a duty to maximize profits, and thus returns, for shareholders. But focusing directly on this fiduciary duty, asking oneself “How can our company maximize its financial outcome?” often bypasses a series of questions that are actually essential to achieving that outcome. These questions revolve around what I call a meta-experiment.
I’ve spent the last decade focused primarily on biotechs - first leading Recursion’s data science efforts (and later engineering, product management, and many cross-disciplinary initiatives) through multiple rounds of funding, growth phases and milestone achievements, and now advising more biotech companies than there were employees at Recursion when I joined (there were 8 of us then). As a result of this, I’ve lived in a world that revolves around experiments. I’m talking about cells in a dish or mice in a cage, genetic knockouts and potential therapeutic treatments, target discovery and chemical optimization. But that experimental world soon extended beyond the wet-lab to affect how I view many other things. I view product development in terms of experimentation (you have a hypothesis of what might be valuable to a potential customer and you test that hypothesis through an MVP implementation). I view organizational design in a similar way (you have a hypothesis of how you might organize your teams and efforts to create greater value, and you test that hypothesis). Same with operating plans, processes, and more. And with a level of abstraction, I view a business as a vehicle to conduct large-scale, complex experiments.
Any good company is ultimately defined by a central thesis - a hypothesis, if you will - that declares what it uniquely believes about the world, and how the world could and should be different. Perhaps it is that tech-enabled commoditization of passenger-driver connection could revolutionize transportation (Lyft, Uber), reusability of rocket components can dramatically reduce the costs of space travel (SpaceX) or that natural products with ethnobotanical priors will have a higher probability of successful clinical translation (Enveda). Each central thesis could be right, or it could be wrong. And a company is simply a vehicle to discover that outcome. Their job is to conduct a meta experiment in the world to test out their hypothesis. With this framing, I see four natural questions required to answer the fundamental fiduciary duty question of “How can our company maximize its financial outcome?”
What is our meta-experiment?
Would a successful outcome of this meta-experiment create meaningful value for the world?
Would a successful outcome of this meta-experiment yield a significant positive financial return for our investors and employees?
Is our team capable of conducting this meta-experiment?
The first question is foundational. Any team running a business that does not have a clearly defined central thesis for their meta-experiment has nothing concrete to focus on, nothing to point itself too to withstand the hardships of building a company. Lack of clarity (people not knowing what the goals are) paralyzes teams and lack of strategic focus (goals that are disconnected, or worse, misaligned) leads to a failure to compound results and build momentum. Without a clear definition of your meta-experiment, your strategy ends up being driven by the interests of others (maybe your employees, maybe business partners). And at that point, it’s not yours - it’s theirs.
The second question is an interesting one, and some might argue it isn’t strictly necessary. For me, I’ve decided that life is too short to focus on problems that a business has, but that don’t actually provide meaningful value for the world. I’ve been there, done that (pre-Recursion days) and have no intention of going back there. But the definition of “meaningful value” is a subjective one, and one could argue that unless a company creates meaningful value for some audience it will never have a financial outcome that can keep its lights on. The opposite is clearly not true though - there are plenty of things that provide meaningful value to the world but do not lead to a positive financial return for investors or employees. And this is precisely why (or at least one good reason) we need academic institutions, government support, non-profit organizations/charities, etc.
The third question is the one that often gets the most focus for a business, but by not separating it from the other three, our assessment can get muddled. Good investors and good leaders need to be able to separate this out, and still candidly assess whether or not they believe a positive outcome to the meta-experiment will yield a significant financial return. I frankly don’t think I’d be very good at answering this question, which is one reason why I don’t think I’d be a very good investor. But separating the other questions from this one actually helps one answer the fiduciary duty question, just like being able to calculate independent and conditional probabilities enables one to determine the joint probability correctly (i.e. P(A and B) = P(A)P(B|A)).
The final question is also equally important. Every leadership team should be looking at themselves and their employees through the lens of “Do we have who we need to successfully conduct the experiment?” and “Is anyone here holding us back from conducting the experiment?” If the team is not capable of executing the meta-experiment, then it’s foolish to even try to conduct it in the first place.
If in your assessment you have a clear answer to the first question, and the answers to all of the remaining three are “yes”, then you have a recipe for a potentially really successful company - one that could truly improve the world and yield a strong financial return. If not, then you’re ultimately just betting on the team and their ability to magically pull a rabbit out of a hat.
As a business leader, fight to create that clarity about your meta-experiment. Make sure it’s clear, focused, actionable. Make sure you communicate that clarity, and reinforce it through action. Staff your team so that they have the highest probability of conducting the experiment. That’s your job - to make sure you have a clear experiment and that you execute it well.
It is important to understand that I am not advocating that your experimental plan needs to be rigid and set in stone. Your plan should evolve, but your meta-experiment - your hypothesis - should not until you have either proven it true or false. Experimentation is often non-linear and meandering, but it always has a destination in sight. If you find better ways to conduct your experiment, use them. If you find that an approach to proving a component of your hypothesis is no longer viable, find an alternative. But stay focused on your meta-experiment. In the end, if data comes back that significantly disproves your hypothesis (that your central thesis will create value for the world and a positive financial return), don’t start grasping at any straw that can give a financial return. Step back, reassess, identify clearly your new meta-experiment, make sure your stakeholders are on board (that the four questions are answered positively) and then execute on that new experiment. While your fiduciary responsibility is to make decisions in the best interest of your shareholders, one of the best ways to do that is to define and conduct your meta-experiment.