This question already has an answer here:
I am trying to split down user stories for a brand new system. This system will record weight readings from a balance connected to a serial port.
Due to regulatory data integrity requirements - the data must be attributed to the user who produced it. The business cannot possibly use the software without meeting this requirement.
In order to meet these requirements we initially wrote a single story covering both login and data recording but this seemed a little large so we tried to split it as follows:
1) As a lab technician, I would like to be able to record weights electronically without transcription so that I do not make mistakes when recording data.
2) As a QA officer, I need data recorded by users to be attributed to them via their windows credentials so that we can comply with <relevant data integrity guidance>.
Applying INVEST to these new stories gives us some issues. These stories could be implemented independently but are valuable only in conjunction with each other.
- If we build just story 1 it seems not to be valuable, because the business could not possibly accept the system without meeting data integrity regulations, therefore no user can gain value from it.
- If we build just story 2 (a login function), a system you can log in to and do nothing provides no value.
Is this a valid way of splitting or do the stories also need to be independently valuable?