backlog refinement

clarifying system demand

All remote:af teams work to sequenced backlogs that are periodically refined. We make the distinction between ‘sequenced’ and ‘prioritised’ because often priority and sequence differ in practice.

Backlog refinement can occur before or during team planning, but if done separately it must match your planning cadence to ensure that the backlog is ready to support planning. Keep the gap between refinement and planning small to reduce the risk of divergence.

To reduce uncertainty and provide transparency we clarify the team’s understanding of backlog items to ensure shared understanding. This reduces the risk of the team committing to work that isn’t ready or that is poorly understood. Where practical it also enables the team to cross-skill, increasing resilience and flexibility over the longer term.

decomposition of value

The customer or stakeholder demands added to the team’s backlog are often presented as significant pieces of work or don’t contain the right information for the team to begin work. Smaller demands with lacking detail can often be resolved through collaboration with the requestor or using scenario oriented techniques such as Behaviour Driven Development (BDD). Larger pieces of demand, we slice the work into small user stories or tasks using a technique like User Story Mapping so that we are creating small pieces of incremental value that comfortably fit into our delivery cadence (e.g. two weeks).

sequencing the work backlog

In a capacity-constrained product development system, we sequence the backlog according to the cost of delay according to Don Reinhertsen’s Principles of Product Development Flow. Cost of delay’s simplest definition is “the cost of delaying the release of a feature”.

At higher levels of the organisation (e.g. Team of Teams), there is value in spending time carefully sequencing the backlog. At a Team level, we use cruder methods as the transaction cost of detailed estimation and planning exceeds the cost of delaying value delivery:

• We review the backlog source for new demand;
• We identify any dependencies between the work items;

• We force rank the work items into an ordered list, optimising for value and flow.

Planning

Remote working teams preserve a sustainable pace through planning to capacity. Using the data captured from previous cycles enables the team to make informed decisions on the amount of work they can complete in the upcoming timebox. Example metrics that can be used to understand capacity are;

• System Time; time spent working on a backlog item
• Throughput; a number of work items completed per unit of time (Daily/Weekly/Monthly)
• Lead Time; elapsed time between work being requested and completed
• Work In Progress; unfinished work in the team’s system

The team allocates work items to the available capacity:

affirm team confidence

We conclude planning with understanding the team’s confidence to complete the sequenced backlog items. Allowing space for the team to surface hidden concerns and opportunities enable the leaders to pre-emptively provide support. This should include;

• A rating of the team’s confidence in the plan
• A list or perceived concerns or risks to the plan
• Acknowledgement of opportunities to innovate

‍