Thinking through a hard decision: a framework leaders can reuse

Some decisions keep you up at night. Should you take on the big client with difficult terms? Hire ahead of revenue? Drop a service line? Change platforms? These decisions feel hard because the stakes are real and the information is incomplete. You will never remove the uncertainty, but you can use the same steps every time so you think clearly, involve the right people and avoid the most common traps.

Step 1: Is this a one-way or a two-way door?

This idea is widely associated with Jeff Bezos, who wrote about it in Amazon's shareholder letters. Some decisions are like a one-way door: once you walk through, it is very hard or expensive to come back. Others are two-way doors: if it does not work, you can walk back out.

Two-way door (reversible)	One-way door (hard to reverse)
Trying a new meeting format	Signing a long lease
Testing a new price on a few new clients	Selling part of the company
Trialing a new tool for a month	Ending a major partnership
Running a small marketing experiment	Letting a key person go

Two-way doors should be made quickly, often by the people closest to the work. Spending weeks on them wastes time and slows the team. One-way doors deserve the full process below.

Help me choose

How much of this process do you need?

If it doesn't work, can you easily walk back out?

A useful follow-up question: can you turn a one-way door into a two-way door? A trial period, a pilot with one client, a shorter contract or a phased rollout can often make a big decision reversible.

Step 2: Write down the real question

Hard decisions are often hard because the question is fuzzy. Write it down in one sentence. Then ask "why?" once or twice to get to what you actually need to decide.

Fuzzy: "Should we hire a marketing person?"
Clearer: "How do we get a steady flow of qualified leads over the next 12 months, and is a hire the best way to do that?"

The clearer version opens up more options, like an agency, a fractional leader or a better referral program.

Step 3: List at least three options

When people face a hard decision they often frame it as yes or no. Force yourself to list at least three options, including:

Do nothing / keep things as they are. This is always an option, and it has its own risks.
The obvious choice.
A smaller or staged version. The pilot, the trial, the part-time version.
Something different. What would you do if the obvious option were off the table?

Step 4: Decide on your criteria before you compare

Before you look closely at the options, write down what matters, roughly in order. For example:

Cash impact over the next 12 months
Effect on our best clients
How much of my time it takes
How hard it is to reverse
Fit with where we want the business to be in three years

Setting criteria first helps stop you from building a case for the option you already liked. Then go through each option against each criterion. A simple table with notes is enough. You do not need a scoring formula.

Step 5: Run a pre-mortem

A pre-mortem is a technique described by psychologist Gary Klein in Harvard Business Review. Before you commit, imagine it is a year from now and the decision has failed badly. Then ask: what went wrong?

Set the scene. "It's a year from now. We went with option B, and it was a disaster."

Everyone writes down reasons, alone, for a few minutes. Doing this separately helps quieter people speak up and stops the group from anchoring on the first idea.

Share and group the reasons. Look for the ones that come up more than once, or that nobody had thought of.

Decide what to do about the big ones. Change the plan, add a safeguard, set an early warning sign, or pick a different option.

People often find it easier to name risks when the question is "why did this fail?" rather than "what could go wrong?" It also gives permission to raise concerns without looking negative.

Step 6: Get the right input, then decide

Who will be affected? Talk to them, or someone who represents them.
Who has done this before? An advisor, peer or mentor who has made a similar call can save you from obvious mistakes. See When to bring in a coach, an advisor or a fractional executive.
Who decides? Be clear. Getting input is different from deciding by committee.

Set a decision date and stick to it. Waiting for perfect information is itself a decision, and often a costly one.

Step 7: Record the decision and review it later

Write a short note. This can be a few lines in a shared document:

The decision: what you chose
Why: the main reasons and what you knew at the time
What we expect to happen: in measurable terms if possible
Warning signs: what would tell us it is not working
Review date: when we will look back

Then put the review date in your calendar. Looking back helps you separate a good decision with a bad outcome (bad luck) from a bad decision (poor process). Over time, your decisions get better.

Using AI to help you think

AI assistants can be a useful thinking partner here. You can ask one to argue against your preferred option, list risks you have not considered, or play the role of a skeptical client. Remove confidential details first, and remember that it does not know your business the way you do. See Using AI assistants to manage email, calendar and notes.

Common mistakes

Treating every decision like a one-way door. This slows the whole business down.
Treating a one-way door like a two-way door. Big, irreversible calls made in a rush.
Only considering yes or no. Look for the third and fourth options.
Deciding alone when you should not, or by committee when you should not. Be clear on who decides.
Never reviewing. Without a look back, you cannot learn.