Difference between revisions of "Story splitter"

From Pearl Language
Jump to navigation Jump to search
m (Alistair Cockburn)
m (Alistair Cockburn)
 
Line 76: Line 76:
|url=http://alistair.cockburn.us/Elephant+Carpaccio+exercise
|url=http://alistair.cockburn.us/Elephant+Carpaccio+exercise
|site=Alistair Cockburn
|site=Alistair Cockburn
|person=Alistair Cockborn
|person=Alistair Cockburn
|title=Elephant Carpaccio exercise
|title=Elephant Carpaccio exercise
}}
}}

Latest revision as of 17:07, 19 November 2017

…umpf, a big chunk.

✣  ✣  ✣

Get a big thing done.

Why split stories?

  • Reduce uncertainty
  • Early and faster feedback
  • Easier time estimation
  • Faster value to customer
  • Visible work progress—visible flow
  • Clearer responsibility
  • Earlier start of work/code/test/execute
  • Easier distribution across team
  • Easy to debug
  • Easy to test and validate

See small is beautiful

Therefore:

Split it into bite-sized chunks that each deliver value.

✣  ✣  ✣



✣  ✣  ✣

Because small is beautiful, there is a strong preference for small items, as they:

  • enjoy shorter short cycle times;
  • show more progress on the kanban board;
  • take less effort to complete;
  • are easier to understand;
  • are easier to test and accept;
  • provide a predicable and continuous flow that facilitates expectation management; and
  • make “past results are a guarantee for the future” come true, just like yesterday’s weather.

To be specific, in planguage this looks like:

Small Stories
Scale   Average number of days per item.
Meter   Track the number of days per item in a control chart.
Wish   ≤ 2 days
Goal   ≤ 3 days
Must   ≤ 5 days
Now   20 days
Therefore  
  1. Split items until:
    1. the item has exactly one crystal clear acceptance criterium (or scenario) specified in gherkin; and
    2. each feature consists of ≤ 4 user stories; and
    3. each user story has ≤ 6 tasks.
  2. Identify and discuss outliers during or right after the daily standup.

Sources