Prioritizing project work for your team can be an exercise in hair-pulling frustration. You have to account for the work initiating from within the team itself, and you have to weigh that against requests coming in from other teams who rely on you.

As a company organized into networks of cross-functional teams — many of whom have both internal and external customers — we've spent our share of time collectively banging our heads against the proverbial wall whiteboard. Today, we use a prioritization matrix technique developed by some of our R&D leaders. But along the way, we learned five big lessons about project prioritization that are worth sharing.

And yes: by "lessons," I do indeed mean "mistakes." Read on, and try not to repeat them with your team.

Tip: Find full instructions for the cross-team project prioritization matrix in the Atlassian Team Playbook — our free, no-BS guide to working better together.

Lesson 1: In Data We Trust

Cross-team prioritization has been a struggle, even just within the team of teams that builds a single product. Part of the problem is that, historically, priorities were based largely on gut-feel. Not only did that lead to some pretty...umm, passionate discussions, it meant that when priorities were ultimately agreed upon, they were based on a lot of assumptions. (And we all know who gets made an ass of when we assume...)

We found it's better to do some data-gathering homework before meeting to hammer out priorities and build a portfolio of projects for the quarter (or year). What's the scope of the problems we need to tackle? How have these problems negatively affected customers and the business so far? What's the value in solving them?

Project prioritization will always be based on opinion to some extent. But data-informed opinions lead to better decisions.

Lesson 2: Prioritize Work Based on Consistent Criteria

Similar to prioritizing based on gut-feel, we often fell victim to group-think. It was common for team leads to gather in a room and brainstorm project ideas that would accomplish their high-level goals. Then, each person in the room would be allocated five "votes." The most popular ideas would get prioritized, and the rest would be left behind.

Prioritization shouldn’t be a popularity contest. Evaluate all ideas and projects based on the same criteria.

To keep sexy-but-ultimately-ineffective ideas from hogging the limelight, use a consistent set of criteria to weigh the value of all ideas and tasks you're considering. Start by laying out all the factors you could potentially optimize for (e.g., timing, budget, revenue potential, etc.) and deciding which are absolute must-haves. Then use those as your basis for evaluating and prioritizing project work. The "trade-off sliders" technique is clutch here.

Lesson 3: Agree on Goals and Stick to Them

You walk into a prioritization meeting. Everyone starts talking about the things they think are important, but you aren't all driving in the same direction. You end up agreeing to do far more than you have the capacity for. You walk out with a stack-ranked backlog, but you can't articulate why one item adds more value than another.

Sound familiar? Yeah.

The solution is simple: agree on top-level goals up front, and stick to them. (Not easy, necessarily... but simple.) If a project doesn't align with at least one goal, it goes in the "won't do" column. As a bonus, understanding the team's North Star makes it easier for team members to make every-day decisions autonomously down the line.

Tip: Google pioneered a framework for setting team and department-level goals called Objectives and Key Results (OKRs). Download this free template, and find detailed instructions for OKRs here.

Lesson 4: Brainstorm Individually, Prioritize Collaboratively

Not so long ago, it wasn't uncommon for a quarterly prioritization meeting to last 3 hours. Ouch! With the group's top-level goal in mind, they'd brainstorm ways to reach it and prioritize those ideas as they went. It was messy and inefficient and based more on opinion than research or data.

Now we do this part asynchronously. Participants develop ideas individually as pre-work and send them around for feedback. Ideas that won't further the group's goal are culled before the prioritization meeting even starts. As a result, these meetings are down to 30 minutes on average. Boom.

Lesson 5: Prioritize Project Work with a Specific Timeframe in Mind

Saying no is hard when there are so many good ideas floating around. So we used to walk out of prioritization sessions with way too many items on the backlog — some of which were very high-impact, but not all that urgent. It was confusing for our teams, and ultimately, demoralizing.

We learned to prioritize projects with a particular timeframe in mind. It's far easier to stay true to your goals (and capacity!) when saying no means "not this quarter" instead of "not ever." We've found that allocating about 50% of your capacity to must-haves and 25% to nice-to-haves works well. That leaves you with the ability to absorb the hot ones that'll inevitably get thrown your way mid-quarter.

A New Way to Prioritize Projects

So where have we ultimately landed, thanks to these hard-learned lessons? In a pretty good place, actually. Thanks to our team of program managers and engineering leaders, we've converged on a technique that can be used either at the team or department level.

All team leads involved agree on their biggest, most important-est collective goal, then write down all their project ideas on index cards. The cards are then placed on a matrix according to how urgent the project is, and how big its impact is expected to be. Next, asks from other teams are placed on the matrix in the same way. Finally, we roughly gauge the effort involved in each and define zones on the matrix that represent must-haves, nice-to-haves, and won't-do-right-now. (Full instructions here.)

This technique is still in its early days around Atlassian, so it'll probably go through some fine-tuning as more teams adopt it. But don't let that stop you and your team from trying it now. And let us know how it goes! Tweet your team's project prioritization story to @Atlassian with hashtag #TeamPlaybook.

. . .

This article originally appeared on the Atlassian Blog.