On Rewarding Effort or Outcomes
As a parent and organizational leader, I've struggled when to celebrate effort vs outcomes. A lot of parental guidance today focuses on effort - nurturing a child's curiosity, supporting their experimentation, and encouraging 'character' traits like grit. Yet, there is a strong sense that results matter too - grades and educational achievement in particular, but also less measurable outcomes such as being a good person.
In the world of startups, outcomes matter. Startups are 'default dead' in Paul Graham's colorful phrasing, meaning they have to grow quickly in order to survive. In this environment, it is very challenging for a Founder to simply say 'good effort' if the product doesn't ship, customers don't sign up, or revenue doesn't grow.
Yet, there is a strong countervailing force. As a Founder, you know building a company is hard work. It is unpredictable. It requires creativity. Forecasts of any kind are speculative. And people are working exceptionally hard. They believe in the work and want it to succeed, so if it doesn't, are you really going to come down hard on them?
I think about these two contradicting impulses a lot. Which is why I stopped in my tracks when I read the following from the CEO of Google:
“To encourage innovation, be ok with failure and reward effort, not outcomes.” - Sundar Pichai (speaking at Stanford, April 2022)
Could his insight be the unlock? Startups and their investors in the Venture Capital industry certainly claim to be in the business of innovation, yet most founders and VCs (and certainly the LPs that invest in the industry) are not 'ok' with failure. VCs may expect failure in a large number of their investments, and therefore need to build around the expected outcome (and steel themselves mentally), but that is not the same as being 'ok' with failure. If this is more than a trite aphorism, we need to do more work here.
Creative Pursuits
Sundar's insight is about leadership under uncertainty. He is presenting a way to manage individuals and lead an organization where novel ideas are required for success.
The three areas I've been working in lately are are crypto, education, and innovative organizational governance models. In all three areas, I am currently a student, a participant, and an investor. In learning about each area, I must consider my own risk tolerance for time and capital. That's why Sundar's insight sparked a connection for me.
Crypto seems straightforward. There is little consensus on outcomes. Innovation is clearly needed to build a viable and important industry, with positive outcomes for customers, operators, investors and the world. This is a widely accepted view, and all participants should be building with that knowledge.
Consider education. The industry is massive, employing over 14M Americans and receiving over $800B in government spending. Perhaps because of this scale, there is no consensus on the job to be done by formal education, either at the k12 or University level. Yet, there is are many brilliant and dedicated researchers, organizational leaders, teachers and investors all aiming to 'improve' education.
Should education investors and operators be ok with failure and reward effort, or focus on outcomes? Is to consider if there is consensus for what success looks like? If there isn't consensus on what success looks like, creativity and innovation is required.
There is not widespread agreement on the set of skills and experiences an entirely online school should provide for student learners. The mode of instruction is too new. We don't yet understand the limits and capabilities of the model. As a result, no one should have a high degree confidence in 'success' for any specific online school.
That doesn't mean investors and operators shouldn't support such innovation. It just means they should do so recognizing failure is a likely outcome, and build a portfolio approach as a result. The portfolio can be built over time (such as an employee working at a few startups over a career) or through a fund (investing in many indifferent companies). All parties, including the student customers, should manage their risk exposure given the uncertainty. If failure occurs, we should favor appraisals of the effort not the outcome for all involved.
Contrast the challenge of building an online school with teaching an eighth grade English class. Teaching English is incredibly hard. I am not dismissing the enormity of the task to engage, motivate and support two dozen fourteen year-olds through Twain and Fitzgerald. But, is there confusion on what success looks like?
Perhaps. This goes to the heart of the debate about education in America today. On the one hand, it would be straightforward to say 'no', there is no confusion about the expected outcomes from a year of instruction in 8th grade English. There would something about competent writing, reading comprehension, ability to synthesize written texts and provide analysis, etc. In fact, this task has been done already - here is the California Common Core Standards for 8th grade English. In the State of California, in other words, there is precisely zero confusion as to what a successful 8th grade English class should accomplish.
On the other hand, would anyone actually sitting in a 8th grade classroom not consider the teacher's efforts deeply creative? Every day, s/he must decide how to connect and inspire a particularly difficult constituency. There are an infinite number of pathways for that teacher to succeed, even when the outcomes are defined as specifically as in the common core. And, any assessment designed to measure standards put forth as 'success' would miss a wide range of important factors such as student motivation and excitement to learn. Consider two equivalently scoring students, one who leaves 8th grade inspired and excited to continue learning, with confidence and context for her studies, and the other who did a good job memorizing information. Is the outcome really the same?
With the Common Core, we say no. Standards are defined, therefore innovation isn't required and, following Sundar's logic, outcomes should matter. This view seems at the heart the teacher and school accountability movement.
Yet, if teaching is inherently creative, perhaps we should be more hesitant in our rush to assessing outcomes. What would it look like to evaluate teachers on efforts not outcomes? That feels like a radical idea which would not sit well with either education traditionalists or reform advocates. The stakes seem too high, for students and society overall. And measurement would be messy - what does 'high effort' look like? More time in classroom? More preparation? How could that be determined in a non-obtrusive way palatable to the professional community in education?
If there is no consensus in how to succeed in classroom teaching, perhaps the smallest unit of measure in the $800B education industry, there are implications for the whole industry. There is no consensus on how to operate effective primary, secondary or tertiary schools. In fact, there is little consensus on the purpose of education as a whole.
We know we need more innovation at all levels of the education industry. Yet, as many new ideas are explored, the chance of failure for each one increases. That has significant implications for how an industry should operate. To consider what that would mean in education, return to the world of startups for an analogy.
The Startup Ecosystem
What degree of confidence is there on how to build a successful startup? Very little. There are a lot of helpful books on the subject (two of my favorites here and here), but certainly no repeatable set of instructions that can be followed with any level of confidence. It is an inherently creative endeavor.
It follows then that startup Founders should be judged heavily on effort. Does this happen? I don't know. There is certainly a sentiment that negative outcomes don't harm a future prospects for Founders, but we are all human and it seems to me that Founders of failed startups do see diminished status in the ecosystem. There is also a lot of noise, in Kahneman's usage, in any effort assessment. So, outcomes tend to fill the void in data.
When startups fail, investors can be 'ok' with failure in the sense that they have built a portfolio around the expectation that a high number of individual investments will fail, but the portfolio overall should be positive. Constructing that fund of 25-35 individual investments from the thousands of potential investments is a deeply creative pursuit. Should VCs be evaluated on effort, not outcomes? (For more on VC industry, see Sebastian Mallaby's great book here).
Returning to Education & Parenting
If we agree we want to encourage innovation in educati0n, should we design a system that rewards effort not outcomes? What would that mean at the classroom, school and district level? What structures are needed as an industry to support individual actors from probabilistic failure in pursuit of industry wide innovation? I'm not sure, but the question interests me.
Which brings me back to the start of this essay. As a parent, when is failure ok and when is it not? Growing up is a deeply creative pursuit. It is individualized. While there is a lot of advice - it is graduation speech season after all - there is not, and cannot be, a singular path for each different person. Children are pursuing their own life path.
The outcome of this reflection is that I am going to lean much more toward rewarding effort than outcome in my life. The more I look at the world, the more I see innovation and creativity - from teaching 8th grade English and to parenting a ten year old, to building a company or investment portfolio. Outcomes involve a lot of noise. Best to push against the natural tendency to look at outcomes, and consider more inputs and process along the way. I'm not sure where this will lead me over time, but it seems to open up a path for more equanimity and less judgement when I am in a position of power and influence.