How to avoid a putrid backlog

"And here is a view of the backlog..." I uttered in a demo of Jira Software the other week. The two seasoned integration engineers shot each other glances and started to giggle.

"What up? " I curiously asked.

"Well, when we talk about backlog, it often just mean a list of old putrid stuff we never get round to do"

I am a list person. I love making them, often starting the day with a short handwritten list on a piece of paper. Makes me feel both energized and a tad in control I guess. No wonder I fell in love with Jira when I saw it back in 2008.

When I started to grasp the concept of cognitive load I understood that the lists help me offload my brain. Instead of needing to remember I trust my list and free up brain energy for the doing part. If your head keeps you up at night with all the tasks you mustn't forget do to in the morning, a simple trick is to just note it down.

If it only were you and your mate doing the work talking regularly throughout the day would be totally sufficient to get the job done. When you and your mate and 15-50-150-250 other colleagues, partners and customers collaborate to reach the same goal, talking regularly with everyone is not so efficient any more. A centrally stored todo-list help us work asynchronously and manage our collective cognitive load.

When it comes to shared lists of work you need to compliment your list making skills with some list management skills:

1. Classify your work items

Mixing all kinds of work items into one list with a "one size fits noone"-kind of workflow and you'll have a putrid list of 1000 items in no time.

Triaging your work load into different types of work help us define their suitable work item lifecycle, identify what people need to be involved when and slice the work list into different sub-lists or boards to become useful in of different contexts.

As the cynefin framework teaches us; different type of work items require different resolution approaches. Some items are resolved by following best practice checklists, some needs experimenting etc

2. Involve and interact

If the list is just sitting there staring at the two members in the team that are supposed to do the work the list will turn putrid quickly. Why should they spend time keep it "up to date" if no one is consuming the information?

Inviting the necessary people into the process is a good way to ensure motivation to keep content alive and useful for all parties. Sharing one platform also ensures content stays in one place and not diverge into various chats, emails, files and whatnot. It will prevent time leakage around finding and putting together "latest status report" type of content.

Identify what people should be involved for each classification. What are their background, need, expectation, interest, driver, fear, skillset etc? We need to take care of the users. Define views, boards, list for them that makes sense from their point of view.

When people collaborate cross team, department and company we need to clarify what to expect from each party. Trust between humans is built when their behaviour is predictable and expectation is managed. It is simple to state a few rules of engagement, but it is very often overlooked. A tip for more reading on types of teams and types of interaction: Team Topologies.

3. Utilise the issue hierarchies in Jira

A putrid list of work often contain loads of unrefined yet fine grained task items. Short one words / oneliners with very little context.

As you can read in this article product backlog items should be coarse grained at the beginning of their life cycle and then well refined when you get closer to the development, building on the "just in time" principle derived from lean manufacturing.

Use the hierarchy of Epic > Story > Subtask in Jira. Start your work on the parent level, the Epic, fill content enough to be useful for the conversation around customer need, context, scope, benefit, prioritization and planning. You can add "story candidates" as a bullet list here. You can connect the Epic to a Confluence page to evolve your user requirement documentation even further(and version control it). When you're getting close being ready to develop, then break it down the Epic to children storys or tasks and stick those into sprints if you do scrum boards or kanban board.

4. Kill your darlings

Creating work items is easy but making the decision to removing them is hard. When people have angst for getting rid of things, the lists become longer, as does the time it takes to organise and refine the work. I once met a product owner that tried to have control over the exact rank of 2000 storys in a board.

Be prepared to clean up "the messy kitchen drawer". If you dare to close or delete work items you also protect the collective cognitive load. Good ideas will come back.

Learn more about Jira here: https://www.atlassian.com/software/jira