Demo data has proven to be both critical to our ability to release new quality features as well as time consuming to get right.
Demo data has multiple stakeholders (we've learned that it's better to combine testing and demo data into one data set that we refer to as demo data for purposes of maintainability):
Automated testing (excluding unit and integration testing)
As part of the definition of done to show that a feature works
Including interoperability with other systems.
Demos (self-guided and presenting a story)
Demos for conferences
To date we've accomplished:
Creating non-trivial lists of facilities in multiple geographies by multiple programs with varying complexity.
Users, some of which are representations of user-personas.
Random / nonsensical performance related data sets.
Utilizing various tools: JSON database definition and Mockaroo.
What we will accomplish in this epic:
Defining and Improving our demo data in services where it's currently weak for any of the stakeholders listed above.
Expanding the set of tools we use and creating exemplars of how and when to use those tools.
Contributing to the Definition of Done guidelines and a toolset for defining and producing demo-data.