SPIDR - Mike Cohn's Approach to Splitting User Stories

S - Spikes

A spike is a research activity that helps teams gain knowledge when facing uncertainty.

• Make a large story smaller by pulling out a spike, which is a research activity after which the team will know more.
• Sometimes just doing a spike makes the remaining work a manageable size.
• Other times, the new knowledge created by the spike makes it easier to see ways to split the story.

P - Paths

Consider the paths through a story and split each path into its own story.

• Draw a simple flowchart of what happens in a story. Each sequence of steps can be a story.
• Expand one big step of the flowchart into a story.

I - Interfaces

Split a story across multiple interfaces if supporting those interfaces makes the story take significantly longer to develop.

• Split out stories by browser type or version, or by different hardware.
• Consider building a minimal user interface first or leave styling out of an interface initially.

D - Data

Look for ways to split the story based on the type of data that must be supported.

• Can a first story support valid data and a later story add support for invalid data?
• How about frequent types of data and less frequently seen types of data?

R - Rules

Sometimes a story is large because of the business rules, technology standards, or such that must be supported.

• Consider relaxing support for these rules in an initial story.
• Add support for additional rules in subsequent stories.

»Mike Cohn →