Below is a simple mindmap for "Stories"
it in our project. In short we:
Write small, deliverable, user stories ( 3-5 days )
Store them in Jira and on sticky notes
Include in each story a brief summary that highlights what it does,
why its valuable, and several tests
Evolve them from simple reminders, to placeholders for our
understanding, and finally an agreement on when we are "done".
This gives us some advantages:
(1) We are able to be more flexible since the details can be fleshed
out when the time is right
(2) We are able to collaborate and focus on verbal communication
instead of passing on written documents
I'll elaborate a bit on of each of these points.
Writing small, deliverable stories
Finding the "right" size for your stories is somewhat tricky. Since
people have very differnet ways of specifing "size" you are likely to
have to experiment and go with what fits your team. For us, it seems
that we have had most success with stories that a single developer
would use 3-5 days on.
Storing in Jira and on sticky notes
The simplest storage method is just to write things on a sticky note
and place it on your wall. It's quick, easy, and visible. Using some
software to store them can also be quite useful. In an offshore
situation that kind of becomes necessary.
Even if you use some software place to store your stories, I have
found that it's good to keep writing those sticky notes and getting
them up on the wall. This way the stories are always visible to
everybody, and there is also just a "feel good" factor in "physically"
taking them down when you are done.
Include in each story a brief summary that highlights what it does,
why its valuable, and several tests
One of the more tricky things is finding out how much to put into a
story. Again, this depends on how your team feels about it.
Before we started using stories we have some small "specifications".
These where typically a few pages with some text and screenshots.
Switching to user stories can be a slight shock to the system. The
most common question I got was, "Where are the details?".
Well, we do actually have a lot of "details" in our stories. These are
represented by the tests and an a GUI prototype. User stories are not
about masking out details. In order to deliver the story you need to
flesh out these details ( as it becomes necessary ), and in most cases
write them down as tests. In my world, the difference between user
stories, and some kind of "specification" is not really that big. It's
how you get there that counts. It's the collaborative, lightweight
"process" of going from an idea to an understanding of when a story is
"done".
In my experience, one of the biggest challenges is to make sure we
actually do the things necessary to make stories work. If you don't
discuss them properly, sincerly write down the tests, and collaborate
during them whole process, then your stories become more of a problem
than a solution.
Evolve them from simple reminders, to placeholders for our
understanding, and finally an understanding of when we are "done"
The reason why stories work is that they are flexible and demand that
you collaborate. A story starts out just as a simple reminder of a
discussion with the client. It evolves as you discuss it with other
people in the team, and you flesh out some details. In the end it
guides you towards aswering that simple question, "when are we
done?".
Ragnar Birgisson, Mar 07, for Agile-Ukraine Google group,
see the discussion thread
it in our project. In short we:
Write small, deliverable, user stories ( 3-5 days )
Store them in Jira and on sticky notes
Include in each story a brief summary that highlights what it does,
why its valuable, and several tests
Evolve them from simple reminders, to placeholders for our
understanding, and finally an agreement on when we are "done".
This gives us some advantages:
(1) We are able to be more flexible since the details can be fleshed
out when the time is right
(2) We are able to collaborate and focus on verbal communication
instead of passing on written documents
I'll elaborate a bit on of each of these points.
Writing small, deliverable stories
Finding the "right" size for your stories is somewhat tricky. Since
people have very differnet ways of specifing "size" you are likely to
have to experiment and go with what fits your team. For us, it seems
that we have had most success with stories that a single developer
would use 3-5 days on.
Storing in Jira and on sticky notes
The simplest storage method is just to write things on a sticky note
and place it on your wall. It's quick, easy, and visible. Using some
software to store them can also be quite useful. In an offshore
situation that kind of becomes necessary.
Even if you use some software place to store your stories, I have
found that it's good to keep writing those sticky notes and getting
them up on the wall. This way the stories are always visible to
everybody, and there is also just a "feel good" factor in "physically"
taking them down when you are done.
Include in each story a brief summary that highlights what it does,
why its valuable, and several tests
One of the more tricky things is finding out how much to put into a
story. Again, this depends on how your team feels about it.
Before we started using stories we have some small "specifications".
These where typically a few pages with some text and screenshots.
Switching to user stories can be a slight shock to the system. The
most common question I got was, "Where are the details?".
Well, we do actually have a lot of "details" in our stories. These are
represented by the tests and an a GUI prototype. User stories are not
about masking out details. In order to deliver the story you need to
flesh out these details ( as it becomes necessary ), and in most cases
write them down as tests. In my world, the difference between user
stories, and some kind of "specification" is not really that big. It's
how you get there that counts. It's the collaborative, lightweight
"process" of going from an idea to an understanding of when a story is
"done".
In my experience, one of the biggest challenges is to make sure we
actually do the things necessary to make stories work. If you don't
discuss them properly, sincerly write down the tests, and collaborate
during them whole process, then your stories become more of a problem
than a solution.
Evolve them from simple reminders, to placeholders for our
understanding, and finally an understanding of when we are "done"
The reason why stories work is that they are flexible and demand that
you collaborate. A story starts out just as a simple reminder of a
discussion with the client. It evolves as you discuss it with other
people in the team, and you flesh out some details. In the end it
guides you towards aswering that simple question, "when are we
done?".
Ragnar Birgisson, Mar 07, for Agile-Ukraine Google group,
see the discussion thread