Improve User Story Quality

Home » Improve User Story Quality

High quality user stories are essential to a successful software project.

No large software project will be delivered on time and on budget if it starts with poor quality requirements or user stories.

Creating high quality stories is hard.

The wonders of the English language give us near infinite ways of expressing the same thing.   This flexibility also leads to potential differences in understanding.

Creating high quality stories is important.

Many software teams do not realise just how important it is to work from good quality requirements or user stories.

Source of Software Defects

Who, What and Why are most important

The business case and functional need are the most important aspects of good user stories.  By this we mean “Who, needs to do What and Why”.  Who is the user (human or connected system), what is the data handled and moved, and why is what follows the “so that” in a user story.

Better user stories – with text analysis

The quality of user stories is ill-served by automation.  In fact, prior to our release of ScopeMaster, we have been unaware of any tools that can help you improve the quality of your user stories. (Update: IBM has just introduced an add-on for DOORS, called Requirements Quality Assistant  that can help with a few aspects of requirements quality, less than ScopeMaster)

Realtime improvement suggestions

ScopeMaster performs realtime analysis and correlation of the text of user stories to help you improve the suitability of the language to achieve, clear, concise, complete, consistent requirements.  By focussing on Who and What we are able to deliver the most relevant and useful suggestions.

Not only does ScopeMaster examine and analyse the language of each story, but it also cross references every story against all of the others, to detect and highlight inconsistencies, omissions and duplicates, that you can then use to refine your requirements.  The smart interface of ScopeMaster dynamically identifies missing stories and makes it even easer to add them.

Handling Infinite Possibilities

ScopeMaster overcomes the vast range of possible expressions of requirements by using a form of Artificial Intelligence know as Natural Language Processing.  This allows you to express your user stories in terms specific to your industry; the tool requires no prior training.

Create better user stories faster

ScopeMaster scans user stories (or software requirements) for appropriate language that will help you write clearer, concise, complete and unambiguous stories.

Detects potential defects

INVEST – is a commonly used checklist for agile user story quality.

  • Independent
  • Negotiable / Concise *
  • Valuable
  • Estimable *
  • Sized *
  • Testable

*ScopeMaster helps the author find and fix these problems (over 50% of all requirements defects).

As a cross-check, we like to use the following list:

  • Clear (unambiguous)*
  • Complete*
  • Concise *
  • Consistent*
  • Correct
  • Current

Again, we see that over 50% of the categories types are addressed by ScopeMaster’s analysis engine.

In our own tests on over 10,000 user stories gathered from over 40 sources, we found that ScopeMaster exposes 0.3 -0.7 defects per CFP (excluding inconsistencies, which we expose but cannot calculate), whilst the typical observed in industry is just under 1 defect per FP (Capers Jones).

Use ScopeMaster interactively at the beginning of your project to improve the quality of your user stories before design and coding is fully underway.  You can continue to refine the stories throughout the development process.

Start your trial now

Alexander Cowan Refined user story

Alexander proposes steps to achieve a good quality user story, proposing the following as a “refined” user story.

‘As the HR manager, I want to create a screening quiz so that I make sure I’m prepared to use it when I interview job candidates.’

We ran this through ScopeMaster in isolation and it instantly detected the primary functional intent and measured the size as 4 Cosmic Function Points.

Mountain Goat Example

We also analysed the set of 238 user stories published by Mike Cohn

  • Time taken
    • 64 seconds
  • Quality assessment:
    • 54% unambiguous, sized at 629 CFP
    • 46% ambiguous
    • 233 potential omissions
    • 28 potential duplicates
    • Over 20 inconsistencies
  • Sizing / Estimation
    • 1161 CFP total size estimate.


Taming software requirements and bringing certainty to software development.

Interpreting software requirements