It's OK to write a short RFC

Building a successful company comes down to repeatedly making good decisions and learning from the bad ones. Request For Comments (RFCs) are my favorite decision-making tool that, if used correctly, can help you with both.¹

RFCs have been long established among most engineering teams² to improve their technical decisions. But you should use them for all your decisions.

What is an RFC? You write down your thoughts about a decision and then ask other people for asynchronous feedback.

Why ask others for their opinion?

• Different people suffer from different biases and by putting our brains together we can arrive at better decisions than any individual could.³

• You make people feel heard and get their buy-in more easily.
• It makes the decision-making process more transparent.

Why in a written format?

• Good thinking is needed for good decisions and it's easier to do good thinking silently. Good writing is good thinking.
• Forcing you to write down your thoughts is a good rubber duck debugging exercise to help you spot faults in your own thinking.
• Written arguments are easier to dissect and respond to than verbal arguments.⁴

• It evens the playing field for introverts vs extroverts.
• Many rhetorical tricks to convince someone of a faulty argument don't work as well in writing.
• Most decisions that affect multiple people need to be documented anyways (if for nothing else but communicating the decision). A simple heuristic: if you need to communicate a decision to multiple people, you might as well write an RFC.
• One of the largest impediments to learning how to make better decisions is called "resulting": confusing the quality of an outcome with the quality of a decision. Written documents act as a snapshot of the context and thinking at the time of decision making, helping you to learn and improve as you look back.
• It’s asynchronous and shines in a remote setting.⁵

• Writing scales. Meetings don't.

“RFCs slow us down.”

If you’ve seen a long and detailed RFC that lists all possible options of a decision and includes an extensive comparison thereof, you might think to yourself:

This must have taken a lot of time to write!

And for your own next decision, you might think:

We need to move fast. Writing RFCs just slows us down.⁶

And you’re not wrong.

How fast you should move depends on the consequences of the decision.

"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. Type 2 decisions can and should be made quickly by high judgment individuals or small groups." — Jeff Bezos, Letter to Amazon Shareholders (2015)

Higher consequence → move slow

• it’s hard to reverse
• many people affected
• high costs
• the decision is for the long-term
• all options are tradeoffs

Lower consequence → move fast

• it's easily reversible
• few people affected
• lower costs
• the decision repeats
• all your options are good

When should I write an RFC?

There's no hard and fast rule, but basically

• if it's easily reversible, doesn't cost a lot, and only affects a few people → you don’t need an RFC.
• For everything else, an RFC is usually a good idea.

A common misconception is that RFCs always need to be long and well-written. That’s not true. An RFC can be super short. Just write down a sentence or two about the problem that needs a decision, and then put down the recommended solution. Your main goal is to transmit clarity of thought. I've seen great RFCs consisting solely of bullet points.⁷

Only for decisions of high consequence does it make sense to slow down and take more time with the decision process (and write a longer RFC).

Here are two templates to help you get started. I use the short template or no template at all for 90% of my RFCs.

Short RFC template → for decisions of lower consequence

• Problem: What made you write this RFC?
• Solution: What do you propose and why?

Long RFC template → for decisions of higher consequence

• Situation: What are the basic facts and background that bring us to needing a decision?
• Options: What are the potential things we could do?
• Criteria: The ways/criteria that you will use to compare these options. If helpful, state non-goals (criteria that are irrelevant) as well.
• Comparison: The actual nitty-gritty of the comparison.
• Recommendation: Your final recommendation.

Who should write an RFC?

Anyone! RFCs do not imply “design by committee”. Asking other people for their opinion does not mean that you need consensus to make a decision. Notably, the person writing the RFC does not need to be the person making the decision. RfCs are about alignment, not about consensus.

Talk, when emotions run high

When you notice (or anticipate) that emotions start getting involved, it's time to drop the pen and have a meeting to make sure no thoughts get lost in translation. I've seen (and participated in my fair share of) misunderstandings that escalated in long written threads that could have easily been resolved in a short 15 minute conversation. ⁸

Write postmortems, too

A postmortem is another written document—this time written when something goes wrong.⁹

They detail

• the impact of what went wrong
• a factual timeline of what happened
• the root cause¹⁰

• the lessons learned
  • what went wrong
  • what went well
  • where we got lucky

The point of a postmortem is not to assign blame, but to learn from your mistakes. That makes postmortems the yang to RFCs' yin.

• By "sunshining" your mistakes you help others learn vicariously from your bad decisions.
• Sharing your mistakes invites others to do the same and is a great example of creating psychological safety by pretending that it already exists.
• Transparency: postmortems add another feather to your cap—showing it’s more than just a corporate value painted on the wall.¹¹

In sum: RFCs (and postmortems) aren't just for engineers. They're a powerful tool for training your decision-making muscle across your entire organization. Just write down your thoughts and invite asynchronous feedback. Move faster when the consequences are low and slow down when the stakes are high. It's OK to write a short RFC.