Decision making as underwriting
Don't think of it as bottoms-up vs top-down
You don’t make decisions as much as you underwrite them. When you think of them as an exercise in underwriting, you are forced to think about upside and downside and accountability and all the implications of when the chickens come home to roost.
The last 10-15 years of tech (now delightfully known ‘the Zirp years’) gave rise to a practice of radical delegation when it came to decisions. It was common to hear executives proudly affirm that they “pushed decisions to the leaf nodes”.
(At the risk of corporatesplaining this phrase: If the organizational hierarchy is viewed as a “tree” data structure, the leaf nodes are individual contributors. The idea was to have people doing the work make as many decisions as possible, rather than managers up in the hierarchy.)
This (sometimes? often?) led to managers not being close enough to the work to know how to make decisions or which decisions would be good and which would not. They would delegate any decision that landed on their table: it was “up to the team, we trust them”. You did everything you could to not be seen as a micromanager.
But the pendulum swings, as it always does, and we find ourselves at a new place, culturally. Decisions being top-down is celebrated and being in the details is becoming a requirement of leadership roles.
The following twitter threads and links therein are the best way to get a pulse on the current zeitgeist, from heads of some of the best, operationally excellent companies today.
As the pendulum keeps swinging, it will be easy for companies to take this mode to an extreme as well. That extreme can result in organizational bottlenecks single-points-of-failure, low quality decision-making, disempowering individuals, etc. Any of these are deathly for startups.
So where do you stop that miscreant pendulum in its tracks and find a balance? Here’s Tom Gayner, one of the best investors of our times, and a prolific writer / speaker about value investing:
Starting at the 1:35:38 mark, Shane Parrish, the interviewer, with his inimitable knack for digging out the gold, catches on to something Tom had just said — a group making a decision — and asks him to expound on how that works at Markel. Tom’s answer is a masterclass on finding the balance between the two extremes our pendulum above is swinging to and fro. He says:
“I think that’s a very nuanced point. In fact, I do use the phrase: I’m the sole CEO and it’s good to have one throat to choke. People make the decision and there is an individual that will be accountable and identified with the decision. But any individual who would make the decision without using the resources that’s available to them — of their friends, their peers, their colleagues, the data — that is stupid. Let’s try not to be stupid. We’re going to make mistakes but let’s not make stupid mistakes.
While we have people responsible for any given thing, the embedded network of relationships and groups that exist are helpful for people to make decisions and feel supported when they are wrong. Because there are people who were also around the table who might not have had exact personal responsibility for the decision, they know they were part of the process and as such they tend to be forgiving, supportive, helpful and resilient in face of the consequences.”
In particular, he points at two benefits of involving a group of people: you get input you might have not have had otherwise, and the group feels more bought in if consulted, which most importantly leads to resilience in the face of any negative consequences.
As the cultural winds shift to leaders being more involved in the key decisions within their teams, it can be very useful to shift the perspective from making decisions to underwriting decisions. For example, it is easy to answer “who is making this decision” with a collective “we” or “the team”, but ask “who is underwriting the decision” and suddenly you’re forced to have a different conversation, and more naturally lean into the nuances Gayner describes.
At the risk of more Corporatespeak, I did find RACI (Responsible, Accountable, Consulted, Informed) (https://en.wikipedia.org/wiki/Responsibility_assignment_matrix) to be a good tool to understand/clarify the meta-discussion around specific decisions - so that it is clear who needs to participate, and in what capacity. Like any tool, though, mandating its use can lead to time-waste/red-tape.