Posts Tagged ‘Task’

At the 2007 Achieve Summit, Gary Markowitz presented The Achieve Approach. This is — from my point of view as a ScrumMaster — essentially a scalable series of questions that needs to be asked and acted upon for each and every thing that an Achieve Team does, no matter how large or small the chore is. To quote Chuck D from Public Enemy, “here come the drums.”

The Achieve Approach consists of four items:

  1. Discover
  2. Architect
  3. Develop
  4. Deploy (or Launch)

If you look at these items as a series of questions that are to be asked of every Task, Story, Project, Sprint, or other work that we do at Achieve, we have a framework to insure that the Team is thinking through a problem, rather than rushing to Development and missing critical bits of information that later crop up as impediments. Even though it may seem — at first blush — a little ridiculous to ask these questions for every Task that a Team creates for a Product Owner’s Story, the scalability of the Achieve Approach will make this a valuable exercise.

Scalability means that a routine (in this case, examining the four parts of the Achieve Approach) is suited for both large and small applications, and that it is nimble enough to apply equally in micro- and macro-environments. The Approach seems to be built for a Project: these four steps are “phases” or “stages” that are dealt with in order to be able to produce the needed information to move to the next stage or phase. This type of approach will be derided by hard-core Scrum practitioners as “waterfall methodology.” Just the fact that you have to complete Discovery before you move on to Architecture sends CSMs into a foaming-at-the-mouth frenzy. “This is not Agile! Waterfall is for lumbering oafs! We don’t need no stinking Discovery!” Although there is some concrete value in having a phased approach, Scrum itself loathes this sort of Six Sigma PM thinking because of the perceived waste of time that comes from doing classical due diligence with rigid phases.

That is not what the Achieve Approach is about. The scalability of the Approach is what keeps it Agile. At a Project level, there may be specific phases that Achieve walks through in order to insure that when we hit the Development phase, we have all of the information needed to successfully Develop. At the Scrum level, it is much more of an Agile process where we simply ask the four questions on a per-Story or per-Task basis in order to bubble up impediments before they interrupt the Sprint. Ingraining these four steps into our everyday thinking should help us ask the questions that need to be asked in order to provide accurate estimates, the correct number of tasks, and bite-sized chunks of work to insure that we can deliver quality and timely releases to our Product Owners within the timebox of the Sprint.

Here’s a real-world example of the Achieve Approach working in a Sprint in question format:

STORY: As a [client developer] I want to be able to [easily theme the four verticals on my website] in order to [implement the new Rotato without making it look like it is a new part of the website]

REQUIREMENTS: [simple functional specifications and use test cases as provided by the Product Owner] — this is to be addressed in a later Art of Scrum

TASK IT OUT: Normally, the Team would start throwing out sticky notes with tasks on them in order to try to meet all of the requirements that were laid out by the Product Owner within the Story; here is where, as a ScrumMaster, I am going to ask for the Achieve Approach to be considered.

  • Scrum it up and bounce out all of the Tasks needed to complete the Story
  • Arrange the Tasks in the order that is needed to complete the Story
  • Keep in mind the Team’s resources, and if Tasks can be done in parallel — this is important for Agility
  • Have the Team consider the four elements of the Achieve Approach: even if it takes 15 minutes extra per phase (mostly Discovery and Architecture I am guessing) it is massively important to think these things through at the Planning Scrum so that estimates are accurate and that the Team does not incur Technological Debt
  • Perhaps each Task can be labeled with the Achieve Approach steps in order to insure none are missed

START SPRINTING: If an impediment occurs, again run through the four elements and see where the impediment was missed, if applicable

DELIVER THE GOODS: Make your Product Owner feel like a Hero by giving them a Product that they cannot wait to demo to their Stakeholders

I would like to see Achieve be able to understand Projects, Stories, and Tasks — especially with their resulting impediments — categorized into Discovery, Architecture, Development, and Deployment so that we could understand better where additional work is being generated by not thoroughly planning ahead of time. Although Scrum is designed to deal quickly with problems as they arise, I still feel that the best way to avoid problems is to think them through in the first place. Perhaps we as a company are not doing enough Discovery. It may be that we are arrogantly trying to Architect on the fly. It could be a common misconception that Launching (or Deploying) is a simple push-button process.

This is a rather long blog (hooray for Nyquil), but the upshot is this: learn it, love it, repeat it: Discover, Architect, Develop, Deploy. This mantra will become a saving grace as we find it reminding us of the path to quality code and products.

Originally published: Tuesday, 12/11/2007

Really good Product Owners always start Planning Scrums with something along the lines of “Once upon a time…” The reason behind this is that one of the least understood but most powerful concepts in Scrum methodology (along with the Squeaky Toy) is the Story. A Story, essentially, is a short description of the usefulness of a feature or product, including vital information such as who is going to use it, why they want it, and what it is supposed to do.

From this trite-sounding invention, you can deduce what is important to who in the Scrum process, the business value, the importance in the grand scheme of the entire project, potential use / test cases, and a whole host of other information within the bounds of a properly-formatted sticky note. Product Owners are the conduit to the customer, or Stakeholders, and as such, they are responsible for translating important pieces of the Product Backlog into Stories. Stories are a bite-sized translation of a feature or client request into terms that the Team can understand, cut up into Tasks, and agree to deliver at the end of the Sprint.

According to Mike Cohn, a properly formatted story is thus:

"As a <some role> I want <something> so that <some value / justification>"

This deceptively simple formula drives Product Owners crazy, because it forces them to get rid of all of the ballyhoo and explain the basic purpose of this feature to the Team. In response, the Team is able to meet the Product Owner halfway and start from the bottom to build a feature or product that the client really wants by describing a series of Tasks that will reach this goal — demonstrable and working — by the end of the Sprint.

Once the Story is told, the Product Owner, the ScrumMaster, and the Team complete the Story by negotiating an Agreement. The Story is usually the top half of the sticky note; the Agreement is the bottom half. In order for the Team to be successful, it must meet the requirements set forth in the Agreement. Agreements are usually a series of bullet points that describe what should be demonstrable to the client by the end of the Sprint.

To put this into perspective, here is a Story with an Agreement from Achieve’s history. Note that Achieve has exercised the First Rule of Scrum: if the rules don’t work for you, throw them out.

STORY:
“This Story allows [the site administrator] to deploy quiz question creation to site users while guiding them to provide these questions in the right format.”

AGREEMENT:
* Exactly four answer boxes, all required
* No title field for users
* No feedback per question
* Auto-generate title field based on taxonomy, question field, and an incremental random number (to prevent repeats in any possible case)
* administrative controls to enable / disable functionality
* show mockups for the way this feature would look to a user

With a Story defined with an Agreement, as long as the Sprint is a reasonable length of time for the Team to perform this work without impediments, there will be a happy ending!