How We Make Decisions

“Horizontal practices” doesn’t mean everyone does whatever they want or that discussions keep going on and on until getting to consensus. Try to think of decisions as bounded experiments that we design to the best of our ability at any given time, so that they are safe to try. Some experiments turn out how we hoped, while others don’t. Either way, we learn and iterate, and we keep everyone on the same page so we can keep moving fast.

TLDR

  1. If the decision doesn’t affect the work of others, is wholly within the bounds of our current strategy, and it’s easy to undo or change directions later, then it’s your call.

  2. If the decision affects the work of others, falls outside our current strategy, or is hard to undo, use the advice process.

  3. If in doubt about which bucket it falls into, ask people around you to sanity-check.

What “Safe to Try” Means

“Safe to try” is not the same as “disagree and commit,” where someone decides (maybe by majority?) that there has been enough debate and the decision will go forward, even though some people still aren’t on board. In “safe to try,” we actually do achieve consensus -- by changing the scope, complexity, duration, mitigants etc., of the experiment so that everyone can agree there’s no reason not to do it.

Advice Process

The advice process is meant to provide procedural clarity so that we can gather the inputs necessary to make good decisions while still moving fast. Here’s how it works:

Roles

  • Proposer(s) defines the problem, proposes a solution, and convenes the advice process. They take on the responsibility of getting the experiment ratified. If there are consent stakeholders, the Proposer iterates until all Consent Stakeholders agree it’s safe to try. If there are only Advice Stakeholders, the Proposer incorporates their inputs and makes a final experiment known.

  • Consent stakeholders are people who get a veto on the proposed solution until it can be modified or rescoped so that it all consent stakeholders agree it’s safe to try.

  • Advice stakeholders are people who have valuable insights to contribute, but whose own responsibilities aren’t much affected by the decision at hand. They provide advice, but don’t block.

Assignment of Roles

The Proposer may include a preliminary list of Consent and Advice Stakeholders in the proposal; however, everyone can elect to put themselves in the Consent Stakeholder or Advice Stakeholder roles during the process as they learn more about the experiment and its implications.

Proposal Contents

Proposals should be written down (but doesn’t need to be a graduate thesis!) and should cover these sections:

  1. Roles: Advice and Consent Stakeholders

  2. Problem statement: What is the condition or opportunity that creates the need for this experiment?

  3. Goals: What does the solution need to accomplish? Be as specific as possible.

  4. Proposal

  5. What is the proposed action / the experiment to be run?

  6. For how long? When will its success be evaluated or the experiment revised?

  7. What risks are there and how are these mitigated?

  8. Context and Background: What information do stakeholders need to know to meaningfully participate?

Cycles: The 3 Ds

  1. Discovery: Ask questions to understand.

  2. Debate: Daylight risks and concerns; consider whether there are alternatives, if appropriate; identify the questions that must be answered to be able to get to a “safe to try” state.

  3. Decision: Final “safe to try” revision of the Proposal is ratified by each individual consent stakeholder, and/or the Proposer presents the (written) final proposal to advice stakeholders

Keep It Fast & Simple!

The whole advice process might take just one meeting, multiple meetings, or no meetings at all (conducted asynch). As long as everyone is satisfied that they understand the Proposal and are able to give meaningful input as Consent or Advice Stakeholders and the Proposer has what they need to take cation, we’re good. No need to have more process than necessary to get there.