Judgment Playbook

Two-way / one-way door

The harder a choice is to undo, the more evidence you gather and the slower you move.

When you need itYou are agonising over a small bet or rushing a big one5 minute readtaught in course chapter 8
Execute and learnManage uncertainty

What it is

Some doors open both ways. Others lock behind you. A two-way door is a decision you can undo once you have learned something: you walk back out. A one-way door is one you cannot undo, because you have spent money, signed contracts, used up trust or changed what others do next. The harder a choice is to undo, the more evidence and discussion it needs, whatever its size.

Where it comes from

The doors come from Jeff Bezos, Amazon's founder, in his letter to shareholders for 2015, published in April 2016. He wrote that "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." Most decisions, he went on, "are changeable, reversible – they're two-way doors" and "can and should be made quickly by high judgment individuals or small groups." He called the first kind Type 1 and the second Type 2.

The economics behind it is older and has no single author. In 1974 the economists Kenneth Arrow and Anthony Fisher argued that when harm cannot be undone, uncertainty alone is a reason to wait for information. They called the value of keeping a choice open "quasi-option value". In 1994 Avinash Dixit and Robert Pindyck built this into the standard theory of irreversible investment. The economists gave the reason. Bezos gave the idea a name a team can use at a Monday meeting.

What it corrects

A large company settles on one way of making every decision. The way that survives protects people from blame: slow, many people consulted, lots of evidence. Then the company uses it even for choices it could test in two weeks and undo in a day. No report shows the cost: the experiments nobody ran. More care cannot fix this, because too much care caused it. In a footnote, Bezos says companies that by habit make Type 1 decisions the light, fast way "go extinct before they get large." So the large firms that survive tend to be over-careful.

Young, fast companies make the opposite mistake. They are proud of moving fast, so they also rush through doors that lock behind them. Bezos says this habit is what stops them growing.

How it works

  1. Write the decision as the first concrete step, not as the strategy behind it.
  2. Split that step into the part you can undo and the part you cannot.
  3. Work out what undoing each part would cost: weeks, money, and whose agreement you need. If you cannot describe a way back, that part is one-way.
  4. Make the one-way part as small as you can while still learning what you need.
  5. Then decide each part in its own way. For the part you can undo, use a small group and move fast. For the one-way part, take time and hear out people who disagree.
  6. Say what evidence would later justify the one-way part, and the date you will check for it.

Worked example

In 2011 Netflix made two decisions usually remembered as one. In July it split DVD-by-mail and streaming into two plans at $7.99 each. Its shareholder letter of 25 July 2011 said what this meant. For members who wanted both, the change was "as much as a 60% increase for them from $9.99 to $15.98." In September Netflix announced Qwikster: the DVD service under a new name, with its own website and billing. On 10 October, about three weeks later, Netflix dropped Qwikster and kept one site.

The lesson is in telling the two apart. The new name was a two-way door. A name, a website and a DVD queue could all be undone, and Netflix undid them. The price rise was a one-way door. The letter of 24 October 2011 treated them very differently. The rebrand "caused a temporary cancellation surge" whose impact was "relatively minor." The price change had "greatly upset many domestic Netflix members" and had "hurt our hard-earned reputation, and stalled our domestic growth." US subscribers fell from 24.59 million to 23.79 million. It was the first fall in the two years of quarterly figures in the letter. Netflix told investors it was "done with pricing changes." It undid the cheap choice within three weeks. The expensive one could not be undone.

In a Business Case Weekly case

The Sega case stops in January 2001. Sega's Dreamcast console is selling well after a price cut, and the PlayStation 2 is in the shops. The decision is how long to keep making the Dreamcast. You cannot settle that by asking which door it is, but you can split it in two. One more quarter of production is a cost Sega can stop the next quarter. Announcing the end of Sega consoles is different. A platform is made of game developers, shop space and places in the release schedule. They all go elsewhere. Sega cannot get them back by changing its mind later. Price each part separately before you argue for either.

In your answer

  • "We could undo … in … weeks, for about …"
  • "The part that locks behind us is …, because it changes … for people outside the company."
  • "So I would move fast on … and wait on … until … is true."
  • "If I am wrong, the way back is … It needs … to agree, and it costs …"

Common misuse

The phrase becomes an excuse to rush. People say "it's a two-way door" about a step that can be undone on paper but not in real life. Think of a trial run that a customer will plan their product around. Or a price test after which buyers remember what your product is worth. Do not assume a step can be undone. Ask three questions out loud. Who has to agree to undo it? How long will that take? What will it cost? If you cannot describe the way back, it is a one-way door, whatever people call it.

References

Practise it on a case

More on thisWikipedia explainer

Fragility audit

Next method

Find the one dependency that could ruin your plan, without trying to predict what will go wrong.

Open the next method