My Approach to Working Agreements

You probably don’t look at your Working Agreement (“WA”) if you have one. Your team put in the effort, and it’s not fully serving you – what a waste! OR… Your team has never put in the effort, so y’ain’t got one – what an opportunity!

Look… I’m not going to sell you on why having one of these is a useful idea. The following is how I use 45 minutes to get 5-9 agreements a team of 5-9 people can start with, and then we iterate, resulting in a living document.

1: Cut a Hole in a Box

(Couldn’t help it.)

Remember 2019 BC (…Before Coronavirus)?

‘Twas a simpler time, and ’twas easier to gather folks around a whiteboard. I mean, we still could, but my local stationery store, 6 months into this pandemic, is yet to sell 6-foot-long Sharpies. Sad.

Now in these Work-From-Home-ier times, usage of online whiteboards for concurrent collaboration is more of a thing. I use Mural. Your team should use… the same tool. And gather around the same virtual whiteboard. Virtually.

2: Put your Junk in that Box


On a blank slate, I ask folks to take 5 minutes to write up ideas for what they individually would want to see on their WA, 1 idea per stickie, a smaller number of stickies per person for larger teams, and a larger number of stickies per person for smaller teams.

The aim is to balance healthy discussion and the time box via a rich yet not unruly number of stickies.

3: Make her Open the Box

(Hmm… that 2006 lyric did not age well…)

I then read each stickie aloud, pausing after each one, giving folks an opportunity to ask for clarification. The idea is not to arm wrestle over how important or deserving a particular idea stickie is, but rather to get shared understanding of it. As facilitator, this is where I find myself cutting people off – not fun, but hey, there’s a slightly tougher conversation to get to…

Oh, and along the way, we group similar stickies.

And That’s

Now it’s time to vote for ideas that they individually would want to see on their WA, a smaller number of votes per person for larger teams, and a larger number of votes per person for smaller teams.

Fun tip: it’s OK here to vote more than once for an idea.

Pro tip: for groups of similar stickies, agree which stickie per group will get the votes for the group, e.g., “the left-most” – this will make it easier to assess voting results.

In Mural, there’s a voting capability, emulating the placing of little sticky dots on idea stickies, and only after the voting session is declared done do you see all the votes, but not who voted for what: anonymous voting. This is WAY better than an in-person voting session for a team name I facilitated, where popular names only became MORE popular, and that’s how you end up with Team “Monkey Knife Fight”. Fun times.

The Way

Now done voting, I drag the popular stickie groups over to another area of the virtual whiteboard, away from the fray, for a focused round of discussion. This is where I use my facilitator’s license to set a threshold of votes an idea needs to have for it to be deemed relatively popular, erring on the smaller side for the number of ideas making the “popular” cut.

The aim is to balance healthy discussion and the time box via a representative yet small number of ideas.

The following discussion is similar to how I end Spring Planning sessions, where we look at the proposed Spring Backlog, and I ask each person individually if they think we can pull it off in the upcoming Sprint. I ask for a thumbs-up or thumbs-down, but if I have time, I’ll ask for the confidence vote to be shown via “Fist of Five“.

Here, I underscore that these highest-voted ideas represent what the team feels should be in their WA. I hold space for another round of clarification, and then I ask each person individually if they think we can commit to this as a start of a WA, recognizing we will revisit this later.

Definitely leave time to dive into why anybody would say “no” to this last question. If an idea seems too controversial, then pull it out, and get commitment on an even smaller set of ideas.

The aim is to get something the team can reasonably feel they can own.

You Do It!

Does the team have a communications hub for artifacts, like their own wiki space? Put the WA there.

Does the team have an invitation series for their Daily Scrum? Put the WA there.

Does the team have a Jira board with an unused Quick Filters area? Summarize each point into cute & team-specific shorthand, then squeeze the WA up there as link-seeming text with no associated JQL: you can click on them, but they won’t do anything. (Once you get a DefinitionOfReady, throw that up there instead.)

For 2 or 3 Retrospectives, ask about WA-adherence.

In a quarter, hold a session to revisit the WA, seriously judging its utility, and remaining open to iterating.

The aim is for 5-9 agreements, preferring the smaller end, since it’s easier to refer to a smaller list colloquially instead of feeling the need to dig out a document.

So there you have it: a living document that’s relatively quick to create, to then start testing out how it serves you, with an eye on iterating.

It takes a little wrangling to settle on something in 45 minutes, but it’s doable, and the speed reflects how the utility of the WA is not in its discussion, but in its practice, so focus on getting to something, knowing you can iterate.

Oh, and if this is for a team that’s rather skeptical of “Agile”, then firstly, they should be skeptical: Agility’s gotta earn its trust (and it will). Secondly, point out how the Working Agreement itself is being developed in an Agile manner before their very eyes: there’s a focus on team, finishing, value, incrementality, feedback, and iterating.