One-Way and Two-Way Doors

Effective decision making starts with understanding, in the long-term, very, very few things actually matter. The vast majority of the decisions made day-to-day are either minutia or easily reversible and can be made quickly.

However, a small number of things (about 1 in 10) matter a lot (in the long term) and are worthy of serious pondering, discussion, investigation, investment, and decision making. Back when I was at Microsoft, a mentor introduced me to the pithy phrase 90% of the decisions you make don’t matter. When I got to Amazon I discovered another way of saying the same thing: One-way and Two-Way Doors.

Instead of articulating this in my own words to help you get your head around this fundamental leadership concept, I’m just going to quote Jeff Bezos from the 2015 and 2016 Amazon Shareholder Letters (which, BTW, are fantastic reads):

“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.” 

“But most decisions aren’t like that – they are changeable, reversible – they’re two-way doors. If you’ve made a suboptimal two-way door decision, you don’t have to live with the consequences for that long. You can reopen the door and go back through. These decisions can and should be made quickly by high judgment individuals or small groups.”

“As organizations get larger, there seems to be a tendency to use the heavy-weight decision-making process on most decisions…The end result of this is slowness, unthoughtful risk aversion, failure to experiment sufficiently, and consequently diminished invention.”

 “Most decisions should probably be made with somewhere around 70% of the information you wish you had. If you wait for 90%, in most cases, you’re probably being slow. Plus, either way, you need to be good at quickly recognizing and correcting bad decisions. If you’re good at course correcting, being wrong may be less costly than you think, whereas being slow is going to be expensive for sure.”

You can tell a one-way vs. two-way door decision by answering these questions: 

“What if you’re wrong? Can you go back to the previous state at all or go back without serious consequences?”

Examples of one-way door decisions:

  • Selecting a processor architecture for the next generation of Control4 controllers and touch screens. Deciding on the processor and chip set and associated tool chain is not reversible without a complete reset. 
  • Making a go-no-go decision on shipping a new product under development, e.g. we recently did this with OS 3. Once you ship a product to customers, you can’t take it back. If your specification of Minimum Lovable Product (MLP) is isn’t lovable enough, or you haven’t fixed the right bugs and you still launch you’re screwed. 
  • Hiring someone full-time. 

Examples of two-way door decisions:

  • Picking the default background for the Control4 OS 3 customer experience. As long as the software architecture enables us changing it later (see question 3 below), this is easily reversible with a software update. 
  • Hiring a contractor (with a clear, time-bound, statement of work and strong supervision, of course).

Here are some questions leaders should regularly ask:

  1. Are we being intentional in identifying which decisions are one-way and which are two-way?
  2. Are we using all tools available to us for driving clarity of thought in making one-way doors? See my blog post on some of my favorite tools here
  3. Are we inventing ways to turn one-way door decisions into two-way door decisions? For example, inventing ways to get closer to the continuous integration/deployment (CI/CD) ideal in product development (see the 2nd example of a one-way door above).
  4. Do the right owners have the autonomy they need to make two-way door decisions?
  5. Are we actively looking for and identifying and correcting bad decisions? Do we ask the 5-whys and then engineer something that corrects the underlying error (e.g. Control4’s Engineering of Error Corrections (EEC)s or Amazon’s COEs)?
  6. Are we applying positive re-enforcement to teams to encourage autonomy and ownership of two-way door decisions?

Relevant Amazon Leadership Principles:

  • Invent and Simplify
  • Are Right, A Lot
  • Deliver Results
  • Think big

More reading:

© Charlie Kindel. All Rights Reserved.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.