In the Scrum method of agile software development, each user story is assigned a corresponding effort estimate by the team that will complete the work.
But what happens when a team can't settle on an appropriate estimate? What if a story includes too many variables to really know how big it is? Or what if its requirements are known, but its effort is off the charts? Such stories are called "epics.
" While a typical story is expected to be completed in four to sixteen hours, an epic usually refers to a story that would require twelve - or many more - hours to complete.
Most Scrum experts recommend that, if a story's tasks will require 12 or more hours, it should be broken down - or decomposed - into its constituent stories.
These decomposed stories will be smaller, more narrowly defined.
In essence, this practice of decomposing epics simply helps a development team translate its work into manageable chunks of work.
But what's the danger of estimating an epic? A best guess can't hurt, right? Actually, estimating epics is potentially harmful because it deludes a Product Owner into the belief that the requirements, tasks, and effort of the epic are known.
Let me explain.
When a team estimates an epic, that estimation is seldom limited to a kind of best-guess reference for the team.
More commonly, that estimate is used as the basis for forecasting, which, in turn, forms the basis of a budget.
At that stage, that best guess has suddenly become an inflexible projection that commits a team to perform an unknown amount of work within a fixed budget.
This approach is like going to the grocery store with a set amount of money to spend, but no list of what to buy.
Clearly, someone in that scenario would have a lot of questions.
What dish am I preparing? What ingredients does it include? And, if I can't budget for all of those ingredients, which ones are most essential? Basically, this hypothetical shopper has no idea how to accomplish the task at hand.
The same goes for the Product Owner - who commits to an estimate of a story when there is little information surrounding the project's requirements.
But what happens when a team can't settle on an appropriate estimate? What if a story includes too many variables to really know how big it is? Or what if its requirements are known, but its effort is off the charts? Such stories are called "epics.
" While a typical story is expected to be completed in four to sixteen hours, an epic usually refers to a story that would require twelve - or many more - hours to complete.
Most Scrum experts recommend that, if a story's tasks will require 12 or more hours, it should be broken down - or decomposed - into its constituent stories.
These decomposed stories will be smaller, more narrowly defined.
In essence, this practice of decomposing epics simply helps a development team translate its work into manageable chunks of work.
But what's the danger of estimating an epic? A best guess can't hurt, right? Actually, estimating epics is potentially harmful because it deludes a Product Owner into the belief that the requirements, tasks, and effort of the epic are known.
Let me explain.
When a team estimates an epic, that estimation is seldom limited to a kind of best-guess reference for the team.
More commonly, that estimate is used as the basis for forecasting, which, in turn, forms the basis of a budget.
At that stage, that best guess has suddenly become an inflexible projection that commits a team to perform an unknown amount of work within a fixed budget.
This approach is like going to the grocery store with a set amount of money to spend, but no list of what to buy.
Clearly, someone in that scenario would have a lot of questions.
What dish am I preparing? What ingredients does it include? And, if I can't budget for all of those ingredients, which ones are most essential? Basically, this hypothetical shopper has no idea how to accomplish the task at hand.
The same goes for the Product Owner - who commits to an estimate of a story when there is little information surrounding the project's requirements.
Source...