BIRD is a framework for clarifying what role people are playing in making a decision. It is similar in function to a RACI chart.

💡 Note: BIRD is not for assigning responsibility for completing work. It's normally best to assign one person to be responsible for each task/goal/project.

Before BIRD

Clarify what the decision is. There might be several related decisions with different BIRDs. Tease them apart.

BIRD: Bound, Input, Recommend, Decide

🚧 Bound

This person says “Do anything, as long as you don’t do X”. They put fences around cliffs to stop us falling off them.

Boundaries shouldn’t be too restrictive - there should still be a lot of space to play in. They might  be “don’t break this law”, or “don’t say this - it’ll cause a PR crisis”.

There also shouldn’t be many people setting bounds (probably 0-2). When you ask someone to bound a decision, specify how you want them to bound it - e.g. “Josh, please can you tell us whether there’s any legislation we need to comply with when we’re setting EAG ticket prices?”

🗳 Input

This person says “I think that option A is most promising - I think option B has downside Y”. They give you information to help you find the high points, the hills and mountains.

There will often be multiple Is. Include people who will be affected by the decision (e.g. people who will carry out the work), as well as people with special knowledge.

✍ Recommend

This person says “I recommend that we pick option A”. They incorporate the information from the Bs and the Is, and make a holistic non-binding suggestion about which mountain to climb.

There is only one R. The R needs to be someone who has broad context, and who is trusted by everyone else involved.

📌 Decide

This person says “we’re going to carry out option X”. They look at the R’s recommendation, and choose which hill we’re going to climb. Once the D has made a call, we’re committed to action.

Ideally, there is only one D. The D should be the person closest to the issue area, with the ability to understand the trade-offs at play. The D and the R can be the same person.

In exceptional circumstances, a group will all be Ds. In this case, there should be a very clearly defined decision-procedure (e.g. “everyone has a veto” or “majority vote”), to avoid paralysis.

After the decision

After the decision there are two further steps (which make BIRDIE if you’re into that):

• Inform people about what we’re doing, and (maybe) why
• Execute the action: make sure tasks are in Asana, we know who’s responsible etc.

FAQ

Example please?

Who decides which speakers come to EAG?

I: Julia/Barry/community members (via form); R: Amy; D: Amy.

When should we use this?

When you notice a complex or repeated decision. When you notice that you’re uncertain about decision-making/ownership. Note that for really small decisions you might only need to say e.g. “Amy is deciding”.

Why have a framework?

Often these things are left slightly mushy, and I think that means that can lead to several people making a decision, which means slow, frustrating, risk-averse decision-making.

What's wrong with RACI?

RACI is not very clear about who is making the decision:

• Is it the responsible party? “Responsible” kinda sounds like they’re responsible for it.
• Is it the accountable party? “Accountable” kinda sounds like they’re accountable for it.
• How much buy-in do we need from the people who are consulted? Do they need to agree with the decision?

This can quite easily lead to a situation where all of the R, A, and Cs are having to agree with a decision. I think that this sort of decision-making-by-consensus is slow and inefficient.

Edit: Brenton Mayer points out that maybe RACI is not trying to get clarity on decision-making. Instead, it’s trying to get clarity on who is responsible for doing a particular piece of work. I agree that it works better for this purpose (though I still think that “responsible” and “accountable” is a bit confusing for this).

Where did this come from?

It’s based on Bain’s RAPID, but changing some parts that didn’t make sense to me. (E.g. Kate, who works for Bain, admitted that “Approve” wasn’t a very good description of what the approver did, and the “Ps” in their framework don’t have any role in the decision beyond being Is. This version also has the roles in the right order.)