Amazon’s 2015 Shareholder’s Letter distinguishes between Type 1 and Type 2 decisions.
Some decisions are consequential and irreversible or nearly irreversible — one-way doors — and these decisions must be made methodically, carefully, slowly, with great deliberation and consultation. If you walk through and don’t like what you see on the other side, you can’t get back to where you were before. We can call these Type 1 decisions. But most decisions aren’t like that — they are changeable, reversible — they’re two-way doors. If you’ve made a suboptimal Type 2 decision, you don’t have to live with the consequences for that long. You can reopen the door and go back through.
Sound advice, but it does hinge on being able to correctly characterize decisions, so I came up with a set of Simple Rules that I hope is helpful. Specifically, these are “boundary rules” that help distinguish between two mutually exclusive alternatives:
- Is there near certainty people won’t be harmed?1
- Can you identify a set of stopping rules?
- Does this consume valuable and irreplaceable resources?
If the answer to the first two questions is yes and the answer to the third question is no, then the decision is likely a two-way door.
Is there near certainty people won’t be harmed? In aerospace, poor decisions can definitely harm — and even kill — people. In many other industries the possibility of routine business decisions causing harm may seem far-fetched, but even something as seemingly innocuous as a dating app or a micro-blogging service can become a platform for harassment.
That said, humans do dangerous but worthwhile things all the time. The potential for harm just means it’s prudent to be more considered.
Ultimately, this rule is about making practical distinctions. Updating a document numbering scheme might inadvertently make thing more confusing, but it’s unlikely to harm anyone; in contrast, eliminating a qualification test for a critical piece of single-string hardware on a crewed space vehicle could easily endanger lives.
Can you identify a set of stopping rules? “Stopping rules dictate when to reverse a decision”, so not being able to come up with any is a good sign the decision is a one-way door.
For example, if crossing a desert of unknown extent, a good stopping rule is to consider turning back when the supplies are half-exhausted.
Less dramatic stopping rules are often applied to development programs, which risk cancellation if they miss important milestones, exceed allocated budgets, or encounter unforeseen technical risks. Being able to identify when things have gone sideways, and what can be done about it, is a strong indicator of a two-way door.
Does This Consume Valuable and Irreplaceable Resources? Obviously Type 2 decisions are allowed to consume resources. The opportunity cost of resources used on Option A is an investment that hopefully pays off, and if it doesn’t, then you’re out the time and probably some money.
This rule is focused on the resources that are so precious that once consumed in pursuit of Option A the other options become infeasible future alternatives. For example:
- Cortez didn’t just decide to invade, he scuttled his ships. Given the dearth of European dockyards in 16th-century Mexico the ships were irreplaceable, so his decision made it practically impossible to reverse his decision and retreat to Spain. (This is perhaps a rare example of someone intentionally creating a one-way door as a motivational tactic.)
- Consuming long-lead hardware is a one-way door because by definition it cannot be replaced without major schedule impacts.
- Large capital expenditures (where large is relative to your accessible capital2) have similarly high opportunity costs because there simply won’t be any money left to try something else if we fail.
To practice using these new rules, let’s look at some intentionally tricky cases.
First is a category of decisions that looks like Type 2 but is often Type 1: decisions that grant rights or entitlements. Take for example the Affordable Care Act. Politics aside, it unquestionably granted new healthcare coverage rights, and although the rules for changing laws are the same as for creating them, Congress is not having an easy go at trying to repeal or replace the ACA. That’s because once given, clawing back rights and entitlements becomes harmful (even though it’s just a return to an old status quo). That is, granting new rights didn’t cause harm, but the decision is nonetheless one-way because reversing the decision to remove them does.
Next let’s consider the opposite scenario: an expensive proposition that looks like a one-way door when it isn’t. Consider deciding between two options: (A) a labor-intensive process with high operational costs; (B) up-front investment in automation to reduce those operational costs.
The labor-heavy option is clearly two-way: (1) it does no obvious harm; (2) we can change our minds and enact the automation plan later; (3) the recurring operational expenses are spread in time, so the exposure is limited.
In contrast, the expensive automation initiative looks like a one-way door because it’s a large up-front expense. However, if there are good stopping rules and the capital is only budgeted rather than committed, then Option B is probably reversible for quite some time.3
Ultimately management isn’t about making decisions, but it is (in part) about articulating a framework within which people understand how to make decision for themselves. This type of decision-making delegation is scary for some leaders because the team will definitely make mistakes, but providing simple rules to better identify the one-way doors will help isolate the missteps to decisions that are easily reversed, and save the hand-wringing for when it really matters.
Macabre fact: Replace “people” with “non-combatants” and “harm” with “killed or injured” and this becomes one of the boundary rules used by the United States to authorize drone strikes. Definitely a decision to be made methodically, carefully, slowly, with great deliberation and consultation. ↩
Amazon can write off the Fire Phone and survive, while your start-up might not be able to weather one disastrous hire. ↩
Eventually, the automation path becomes a Type 1 decision. If successful, people’s jobs will change and/or disappear, processes will atrophy, and tribal knowledge will fade. At this point, returning to the way things were will be akin to starting over. The personnel and cultural resources that executed the old process won’t have been consumed, but will be unavailable all the same. ↩