5 mins

A common area where managers fail to scale is in decision-making.

The two extremes are: being responsible means making all decisions and enabling a team means staying out of details and letting them make decisions.

Both of these are bad. One is controlling and fails to scale for obvious reasons. The other fails to scale because it pushes all the work onto the team, creating variance and evading responsibility. Managers in this category will be fine as long as they have high-performing individuals making good decisions, but they won’t know where to begin with people who can’t make good decisions, or how to address issues when things go wrong.

What does a good decision look like?

We can’t A/B test most decisions: they can only be evaluated after the fact, and many other factors may have influenced the observable state. Some prerequisites for a good decision are:

  • It’s a decision, not a non-decision. (Deciding not to decide is a decision, not deciding is a non-decision).
  • It explicitly states the desired outcome.
  • It considers context. For instance, medical decisions are ‘informed decisions’ because we can’t A/B test on health outcomes.
  • It considers what could go wrong (a pre-mortem can be a way to explore this).
  • It states assumptions and provides a way to test them.
  • It states tradeoffs and how they might need to be managed.
  • It is informed by feedback from people with different perspectives.
  • It’s rational and feels possible to those who have to execute on it. (‘We’ve decided to grow by 45% year on year’ is not a decision; it’s a dream.)

How do you get there?

First, what are you deciding (and why)? Determining the actual decision and scope is the first step towards having a rational conversation about whatever it is that is happening; it’s worth talking about the why of the decision and explicitly stating the desired outcome. Failing to do this results in people discussing from different premises, or just not weighing in at all because the scope seems overwhelming. Think about how you can frame the decision as a concrete question, for example, ‘What projects should we invest in in order to meet our growth target of 45%?’

Second, what is the necessary context? This is the landscape the decision is made in and may include: market and team constraints, data points, and previous tests and their results. The point is not to overwhelm people with information, but to make sure that there is a base level understanding of the situation so that everyone can operate from a similar reality, avoiding assumptions and drive-by comments like, ‘We tried X before and it didn’t work’. It’s often worth stating the context around time in engineering: we need to show significant impact over the next quarter vs. we see this as a long-term infrastructure investment, will drive vastly different types of projects. Engineers will often default to the latter, and so when that’s not the case, state it upfront.

Third, fostering discussion. This is the part of the process where you reach a shared understanding and buy-in; where people have an opportunity to gain understanding and ask questions. The decision may include multiple rounds of different kinds of discussion. A pre-mortem can be a specific kind of discussion that is useful to explore what can go wrong before the project even starts. Later on, there might be specific discussions around proposed approaches or impact. Most decisions do not occur in a vacuum – particularly in complex organizations. It’s worth considering who is impacted by the decision and inviting them to share their perspective or concerns.

As part of these discussions, it’s vital that you create space for feedback. Saying you’re open to feedback is easy, creating the space for the right kind of feedback at the right time in the process is much harder. Feedback on the framing of the decision (does the decision make sense? Is there any context missing?) is also helpful. The earlier more structural feedback is shared, the better. Making space for this feedback can include things like:

  • Spacing out the pieces of the decision to give people time to digest, ask questions, and make comments.
  • Facilitating discussion so that a broad range of opinions are heard – including dissenting opinions and asking people who are less inclined to share what they think.
  • Working behind the scenes with people 1:1 to get feedback they are less inclined to share publicly, but ideally, making them comfortable to share that feedback with everyone else.

As you work through discussions and feedback, tradeoffs will emerge that are worth clearly articulating. What do we give up by choosing one thing over another? What do we give up by investing time over choosing the most expedient option? This is often a method of capturing people’s concerns in a way that shows they have been heard and considered, but that other factors are being prioritized right now.

Once it’s time for the decision to be made, ensure the decision is made and communicated. It can be helpful at this stage, especially if it’s liable to be contentious, to communicate what went into the decision, what factors were considered, and who was consulted. Big decisions should often include checkpoints where the decision is revisited, e.g. you might decide to work on a project in support of a goal, but set points where you’ll review whether there’s impact or not.

Final thoughts

Remember that even the best-made decisions sometimes don’t turn out the way you’d hoped. Hindsight is 20:20, so revisit decisions and figure out what you can learn from them. Was there context that was missed? Tradeoffs that turned out to be different than expected? Recognizing a good decision-making process regardless of the outcome is important in order to avoid your team always optimizing for the safest path.

This process might seem arduous, and in some cases arduous is necessary. However, it doesn’t have to be, and many smaller decisions may include these points in a lightweight way. The discussion might take place during a short thread in your chosen messaging app, or even be as succinct as voting by emoji, for example.

As a manager, you’re responsible for the performance of your team – it’s understandable that’s scary. Resist the instinct to control or evade, and build up processes and teammates instead. Empowering your team increases buy-in. Dictators inspire obedience more than enthusiasm. Take some inspiration from my friend Mekka, who when he went on parental leave challenged each of his directs to have made a decision he would have disagreed with. What a great way to help them learn.


With thanks to Michael Norman for suggestions and feedback.