Extra Arm AI
How it worksWhat you getCommunityPartners
How it worksWhat you getCommunityPartnersLog in
Log inStart free
← All posts

Founder playbooks · November 11, 2026 · 7 min read

One-way and two-way door decisions for small business

Short answer: a two-way door decision can be undone, so you should make it quickly and fix it if it is wrong. A one-way door decision is hard or impossible to undo, so it deserves time, facts and care. Most small business decisions are two-way doors, and treating them like one-way doors is one of the main reasons owners move slowly.

Every business owner makes dozens of decisions a week. Which headline to use. What to charge. Whether to sign a contract. Whether to hire. It is tiring, and the tiredness often turns into delay. Decisions pile up, and the business waits.

Jeff Bezos wrote about this problem in his 2015 letter to Amazon shareholders. His answer was not to decide better in general. It was to notice that decisions come in two kinds, and to treat each kind differently.

This guide explains the idea and gives you a simple way to sort your own decisions.

What is a one-way door decision?

In the 2015 letter, Bezos described some decisions as consequential and irreversible, or nearly so. He called them one-way doors. Once you walk through, you cannot easily come back. He wrote that these need to be made methodically, carefully and slowly, with a lot of thought and advice.

For a small business, one-way doors are things like:

  • Signing a long lease or a long supplier contract.
  • Ordering a large amount of stock you cannot return.
  • Taking on a loan or giving away part of the company.
  • Hiring a person, especially a first employee.
  • Choosing a business name you will build a brand on.
  • Anything public that could damage trust if it goes wrong.

These are worth slowing down for. Get advice. Sleep on it. Ask what happens if you are wrong.

What is a two-way door decision?

Bezos called the other kind "two-way doors." These can be changed or reversed. If you walk through and do not like what you find, you can walk back. He argued that these should be made quickly, by one person with good judgment or by a small group.

Most daily decisions in a small business are two-way doors:

  • The words on your homepage.
  • The price of a new offer, at least for the first few customers.
  • Which email to send this week.
  • Which social channel to try first.
  • A small ad test with a fixed budget.
  • The order of tasks on your list.

If one of these goes wrong, you change it next week. The cost is small. The cost of waiting a month to decide is usually bigger.

Why do businesses get this wrong?

Bezos pointed out a trap that big companies fall into. As they grow, they start using the slow, careful process for every decision, including the reversible ones. The result is slowness, too much caution and less invention.

Small businesses fall into the same trap for a different reason: fear. When it is your own money, every decision feels permanent. A new price feels risky. A new headline feels like a big statement. So you wait for more information that never quite arrives.

The fix is to ask one question before you start worrying: if this goes wrong, can I undo it? If yes, decide today.

How much information do you need to decide?

In his 2016 letter, Bezos added a rule of thumb. Most decisions, he wrote, should be made with somewhere around 70 percent of the information you wish you had. Waiting for 90 percent usually means you are too slow.

He paired that with a second skill: being good at spotting and fixing bad decisions quickly. If you can correct course fast, being wrong is cheaper. That is exactly what makes two-way doors safe.

For a small business, 70 percent often means: you have asked a few customers, looked at what competitors do, and you have a reason for your choice. That is enough to try it.

What does "disagree and commit" mean?

The same 2016 letter describes another habit for speed. When a team cannot agree, someone can say they disagree but will commit to the plan anyway. Bezos gave an example where he doubted a project, said so, and backed the team fully.

In a small business, the person you most often disagree with is yourself. You go back and forth between two good options. Disagree and commit means picking one, giving it a fair test, and not reopening the decision every day. Write down when you will review it, and leave it alone until then.

How do you turn a one-way door into a two-way door?

Many big decisions can be made smaller. This is one of the most useful skills an owner can learn.

Rent before you buy. Use a short lease, a shared space or a month to month tool before you sign a long contract.

Test before you stock. Pre-sell a product or order a small first batch. Our guide to testing with almost no money shows how.

Freelance before you hire. Try a short paid project before you offer a job. Our guide on when to hire a human expert covers when to bring someone in.

Launch to a few before you launch to all. Offer a new price or product to ten customers first. If it works, roll it out.

Each of these turns a one-way door into a door you can walk back through. You still learn what you need to, without betting everything.

Who should make two-way door decisions?

Bezos wrote that reversible decisions can be made by one person with good judgment or a small group. In a big company, that means pushing them down to the teams closest to the work. In a small business, it means you should stop being the bottleneck for every small choice.

If you have a helper, a freelancer or an AI team, give them a clear rule. They can make two-way door decisions on their own, inside a budget you set. They bring you the one-way doors, with two or three options and a recommendation. Write the rule down so nobody has to guess.

A simple version:

  • Decide alone: wording, layout, the order of tasks, which customer to contact next.
  • Decide alone within a limit: small spending, such as a test with a fixed cap.
  • Bring to the owner: anything that signs a contract, spends above the limit, changes prices for existing customers, or speaks publicly in the owner's name.

This one page of rules can save you hours a week. It also makes your team faster, because they are not waiting for you on things that do not need you.

How to use this in your business this week

  1. List your open decisions. Write down everything you have been putting off, big and small.
  2. Mark each one. Write "two-way" next to anything you could undo within a month at low cost. Write "one-way" next to the rest.
  3. Clear the two-way list today. Give yourself one hour. Make each decision with what you know now. Write a review date next to each one.
  4. Shrink the one-way list. For each one-way door, ask: is there a smaller, reversible version I could try first? Often there is.
  5. Give the real one-way doors proper time. Book a slot, gather the facts, ask someone you trust, and set a date to decide.

If you use Extra Arm AI, your team already works this way: it drafts posts, emails and options, and you click to approve. You can tell your chief of staff a standing rule: "Treat copy changes and small tests as quick decisions; bring me anything that spends money, signs us up to something, or speaks publicly in my name, with options." The playbook our team follows when it plans uses the same split. You can start free to see it.

This idea pairs well with working backwards: write what the customer gets, then decide fast on everything you can undo.

Questions people ask

What is the difference between Type 1 and Type 2 decisions?

Bezos used both names. Type 1 decisions are the one-way doors: hard or impossible to undo. Type 2 decisions are two-way doors: they can be reversed, so they should be made quickly.

Is pricing a one-way or two-way door?

For a new offer, pricing is usually a two-way door, especially if you start with a small group. Raising prices on existing customers needs more care, because it affects trust.

What if I make a two-way decision and it goes badly?

Undo it, note what you learned, and move on. That is the point of the category. A wrong reversible decision costs less than a month of waiting.

Should a business owner ever delegate one-way door decisions?

Rarely. Delegate the research and the options, but keep the final call on anything that spends a lot, binds you to a contract or carries your name in public.

Sources

  • Amazon 2015 letter to shareholders (Type 1 and Type 2 decisions), Amazon via SEC
  • Amazon 2016 letter to shareholders (70 percent rule, disagree and commit), Amazon
  • An insider look at Amazon's culture and processes, Amazon
Try it on your business

Start from your website or from an idea.Setting up is free.Nothing runs until you say so.

Start from my websiteStart from an idea
Keep reading
  • Billionaire founder habits that work in a small business
  • Work backwards like Amazon: write the announcement first
  • Circle of competence and moats for a small business
Extra Arm AI

Built and run by AI.You make the calls.

Product

  • How it works
  • What you get
  • Start free

Resources

  • Blog
  • Learn
  • Compare

For

  • Use cases
  • Trades

Community

  • Community
  • Partners

Legal

  • Privacy
  • Terms
© 2026 Extra Arm AI