Quit: The Power of Knowing When to Walk Away

by Annie Duke

Instead, he asks them what success would look like over the next few months. And he asks them for specifics. That conversation allows him to sit down with the founder and set performance benchmarks that would signal that the company was heading in the right direction. Then, they agree when to revisit those benchmarks and, if the venture is falling short, to have a serious discussion about shutting it down.

If someone doesn't have kill criteria, this is a way to create them in the midst of failing without losing their identity

There is no honor in spending every last bit of investor money pursuing an endeavor that’s failing. Returning capital to investors is the responsible choice under those circumstances and demonstrates the ability to make the hard decision when it’s the right thing to do.

While this is a concrete example, it seems like it could be a metaphor for other assets, like emotional assets or free time or sleep.

Joining a start-up means working for little pay in exchange for the promise of equity. These talented individuals are willing to make that trade because they believe they are building something world-changing and, if successful, they will reap the benefits of doing so. Once it’s clear that’s not going to happen, continuing on is just trapping them in an endeavor that’s going to fail. It’s stopping them from being able to move on to something better.

Another thought that probably has metaphor potential.

looking to see if the more-optimistic participants had better performance on the math problems or found Waldo more often. While they did find that more-optimistic people stuck to the tasks longer, the optimists failed to perform measurably better on these tasks than the people who were less optimistic. In other words, they quit later, but to no benefit.

The result? Optimism unchecked by realism prevents you from quitting when you ought to walk away.

Optimism is good but must be tempered by realism to more correctly calibrate expected value (EV)

if there is such a switch, the founders are the ones who have to flip it.

this made me think of my "fuck it switch" and this should absolutely be mentioned in relation to this book

you should find at least one person to be your quitting coach. Second, you should try to serve in that role for the people you love.

Deb has been really good at this without any distinctions from this book

this suggests a good strategy for businesses that want to get better at their quitting decisions: When possible, divide and conquer. Have the people who make the decisions to start things be different from the people who make the decisions to stop those things.

this is why changes in leadership are often so drastic but can be good for the future of the business

When the ants first get to a territory, they’re exploring it for opportunities. Once they find an opportunity, a high-quality food source, they start to exploit that. But some percentage of them never stop exploring. That allows them to discover backup plans. Those backup plans are really good to have when the ants are forced to quit, when the ants get unlucky and the food source goes away.

Of we commit 100% to one path without continuing to explore other options, even at a reduced level because you are using some resources for the main objective, you might miss out on you next objective and/or an even better objective. This is an important part of the explore-exploit cycle and where it will be successful long term.

The world might force what you’re pursuing away from you. Or you might be the one who chooses to abandon it when the circumstances of what you’re doing change. You could be in a job you’ve loved with a boss who’s been a mentor to you, but that boss could leave and be replaced by someone toxic.

this has definitely happened to me. this changes the expected value of your current path of action.

When it comes to our aversion to closing accounts in the losses, the pass-fail nature of goals makes this problem worse. As soon as you set a goal or a target, you put yourself immediately in the losses, at least in relation to your distance from the goal.

this is why it is important to reframe goals so that a failure can still be a success.

the pass-fail nature of goals can impede progress, cause escalation of commitment, and stop us from considering the progress we make along the way as a success. The shame in all this is that those finish lines are often arbitrary.

To achieve the things we want to achieve, we have to be responsive to the way the world is changing around us and the way that we ourselves are changing. That would mean unfixing our goals, but we don’t naturally do that.

can a goal be a scaffolding for improvement in an area? As in, the real measure of success (EV) is improvement but the goal gives us a finish line to strive for. And we can revisit when things are evolving to see if we need to adapt or quit the goal.

An “unless” is a powerful thing. Adding a few well-thought-out unlesses to our goals will help us achieve the flexibility that we’re seeking, be more responsive to the changing landscape, and reduce escalation of commitment to losing causes.

A good set of kill criteria means you can win by achieving a goal or by successfully following those kill criteria.

When we worry that quitting means we’ve failed, what exactly are we failing at? If you quit something that’s no longer worth pursuing, that’s not a failure. That’s a success.

We need to start thinking about waste as a forward-looking problem, not a backward-looking one. That means realizing that spending another minute or another dollar or another bit of effort on something that is no longer worthwhile is the real waste.

Goals are pass-fail in nature. You either reach the finish line or you don’t, and progress along the way matters very little. Don’t just measure whether you hit the goal, ask what you have achieved and learned along the way.

key results in OKRs can be deprecated/cancelled. How good would it look for you to have a post mortem with lessons learned for every deprecated goal/key results.