5 Ways to Make Better Decisions With Your Team

Woman holding crystal ball — Photo by Jonathan Sebastiao on Unsplash

I’ve been thinking about how teams can make better decisions after reading Annie Duke’s book How to Decide. It’s a short read with a lot of useful checklists. It’s applicable to both personal and business decisions, but I’ve been thinking most about 5 takeaways that can help your team make better decisions.

1. Be Aware of Outcome & Hindsight Bias

We make decisions without perfect information, so there’s no perfect correlation between decisions and outcomes. There’s always the chance that an “unknown unknown” or just luck can intervene and cause a good decision to lead to a bad result, or a bad decision to lead to a good result. But outcomes are very salient for us, and so it’s easy to succumb to bias and say a decision was good because it had a good result., or a decision was bad because it led to a bad result.

Similarly, hindsight bias distorts our view of the decision, making us think we knew things before the decision that we actually only knew after the fact. One thought experiment Duke gives that stuck with me is this (paraphrasing here):

A) After college you move to a new city for a good job opportunity, not knowing how you’ll be able to tolerate the city’s weather. You end up hating the winters and move back home. How likely is it that someone will say something like “I knew you wouldn’t be able to handle the winters” to you?
B) After college you move to a new city for a good job opportunity, not knowing how you’ll be able to tolerate the city’s weather. You end up loving the job and find out the weather isn’t a big deal for you. How likely is it that someone will say something like “I knew it’d be worth it for you for the role?”

If the answer to A and B is both likely, then there’s probably bias at play -- someone can’t simultaneously know that both outcomes were going to occur.

2. Get Unbiased Input From Your Team & Others

When making and reviewing decisions, gathering more information is extremely helpful. But to do this correctly, you need to make sure the person giving feedback both feels safe in giving it and is not biased by your own perspective.

Psychological Safety. For group psychological safety, you can set norms that emphasize that no one will be punished for sharing information about a decision. For example, you can have something like failure amnesty, making it a norm to review and learn from decisions. In decision-making meetings, you can start by making it clear that everyone is encouraged to share and no one will be punished or criticized for what they say.

For team meetings, I also recommend starting with an activity like a team safety check. This will give you a quick read on how safe people feel sharing, and an opportunity to discuss and improve team culture if needed. I’d repeat this activity until responses are mostly 4s and 5s, and then consider checking regularly (e.g. once a quarter if the team is the same, or once a month if the team is growing quickly) to make sure the safety rating hasn’t changed.

Unbiased Input. Getting more information is only useful if people share their true opinions. If you’re working with a team in real-time, you can do this similar to how you’d do a retrospective and have everyone write down their opinion at the same time and then have people share, making sure that more junior members of the team share their views prior to more senior members. If you’re making a decision like whether to hire someone, you can define and agree on your evaluation matrix beforehand, and then have everyone fill it out independently prior to team discussion.

If you’re seeking information from an individual rather than from a team, Duke recommends mapping out each step of a decision and then asking the person what they think or what they would do. So, for example, first you would describe or show them a job description, and ask what they think. Then you could show them interview questions, and repeat. Then you can show them a potential candidate’s answers, etc. Here it’s most important that you ask the person for their opinion before you share your own, since sharing our own opinion tends to bias people toward repeating that opinion.

3. Conduct Effective Postmortems

The fix for outcome and hindsight bias is to focus on your decision-making process instead of outcomes. To do this, conduct postmortems with your team. Traditionally, I’ve most often seen these done after an unexpected negative outcome, such as the company website going down or a major customer-facing error. After reading How to Decide, I see that there’s also value in doing these for positive outcomes.

In a traditional postmortem, the goal is to understand why and how an outcome occurred and see if there’s a process improvement that can be made to reduce the risk of that outcome happening again. If the postmortem is on a positive event, you can see whether there’s a change you can make to increase the likelihood of it happening again, or if you were experimenting with something that you now want to make permanent.

In identifying why and how an outcome occurred, Dukle also recommends distinguishing between outcomes that are under your control and outcomes that are genuinely due to luck.

Postmortems are most useful when events are surprising. For example, maybe an outcome wasn’t one of the possible scenarios you mapped out before making the decision. Maybe you reached an expected outcome, but there was an unexpected obstacle on the way to that outcome. Or maybe you ended up at an outcome you thought was low probability.

If an outcome was due to something under your control, Duke sees the fix as more knowledge. So you can ask what information could you have gathered beforehand that would have let you anticipate this outcome? Or what unexpected success/failure led to the outcome? How can you better plan for it in the future?

Here I'd also add the fix of better processes. In complex decisions involving multiple people, it’s easy to plan correctly but then execute poorly. If execution is the issue, the fix for negative events is likely a process that’s easier to execute and reduces the risk of error. For positive events, it can be identifying the unexpected positive action someone took, and making it part of a repeatable process.

For outcomes outside of your control, the postmortem can be useful in reminding you to reflect on the decision itself instead of just the outcome. So, for example, if you try a new deployment process and then your cloud service provider crashes, recognize that you shouldn’t give up on the process just because the outcome wasn’t what you wanted. Conversely, if you reached a good outcome only because of luck, you want to make sure you don’t repeat a bad decision just because good luck saved you. So you can ask the above questions to increase the likelihood of a good outcome that’s within your control.

4. Conduct Effective “premortems”

I really like the idea of doing a premortem as well as a postmortem. In a premortem, you ask your team to imagine a negative outcome, and to imagine why it occurred. For example, maybe you set a goal of 20% growth over the next quarter. You could then ask your team to imagine that it’s 3 months later and the team hasn’t met the goal. What factors in your control would be most likely to lead to failure? What about external factors?

Duke notes that research shows that anticipating obstacles leads to better outcomes (and I’ve also seen this mentioned elsewhere). In doing the premortem you shift you and perspective, causing people to think about things you and your team may have otherwise ignored.

5. Lower the Stakes

You might be wondering how to possibly do all of this when there are so many decisions you make every day. Fortunately, Duke explains that one way to be better at decision-making is to learn to identify which decisions are high stakes and need more time and attention, and which can be made quickly.

One way to lower the stakes of the decision is to lower the risk involved. If, for example, you know that any negative impact can be contained or that a decision is reversible, you can move forward more quickly. An example Duke gives is ordering at a restaurant when you know you can come back later - whatever option you don’t choose you can pick the next time.

I’m a big fan of using experiments to lower risk. Experiments let you try out a decision without having to commit to it, e.g. testing a new process for a few weeks or doing a small product test. So you get all the upside of trying out the decision, without the downside of being stuck with a negative outcome. Experiments are also great because they allow you to gain more information, which then allows you to make an even better decision in your next iteration.

---

I hope this helped. If you try any of these techniques, let me know! You can comment below, or email tltcoaches@gmail.com. And if you’re interested in coaching to get feedback on your decisions, you can schedule a free session with me.

Sara JerussMarch 15, 2022