We get a ton of hits on our Agile Approach page so I thought I’d take an opportunity to give some more background on our methods and share some of our experiences.  The Limina Team comes from a traditional Waterfall or Rational Unified Process software development background.  Our consulting history and practice methodology is an adaptation of best practices developed in-house by user experience professionals and collaborators who’s expertise run the full spectrum of user interface; research, analysis, strategy, management, interaction modeling, information architecture, design and development.  As a result, we have developed a suite of services that can be applied throughout the full span of the product development lifecycle.

As a practice, our methodology has been flexible enough to add value to every client engagement in our portfolio.  As our team engaged on increasingly more frequent Agile/SCRUM driven teams, a trend which began for us in 2006, we needed to make some adaptations to keep pace with rapid iterations.  The following is a rough breakdown of the adaptations we made by “activity”.

We tool a look at the activities and deliverables we execute in a more verbose and lengthy cycle and dissected each phase to  determine which tasks and deliverables dovetail with the Project, Incremental Release and Sprint cycles of an agile project life cycle. Each level invariably incorporates tasks and deliverables from the traditional Analysis, Definition, Design, Development phases.  Here’s what we came up with:

Product Cycle: Assuming 6-9 months.

Intense Observation and Analysis activity in the all important “Phase Zero” a period of three to six weeks.  In the absence of the waterfall lead time, this is the cycle where UX research seeks to identify the following in rank order:

• Task / Activity Model
• User Role Model
• Business Stakeholder Goals
• User Goals
• Competitive Analysis

Release planning: “Iteration Zero”.

While rough concepts are being established to determine technical frameworks and baseline use cases, the UX team takes 2 weeks to elaborate on the primary storyboards to cover feature definitions in the first iteration.

The At this phase, requirement gaps have been identified, rudimentary user typologies have been identified, development road map has been established based on technical complexity, feasibility, business benefit and user benefit.  Relevant personas for the user stories and features are drafted, usage scenarios are  drilled down, the draft interaction model is established and the associated process flows and wire frames are generated.  This iterative cycle is a lather, rinse,and repeat.

Ideally after 2 weeks iteration zero kicks off user stories, wire frames and mockups and will be available for itteration one development work.

While UX team is cranking out user stories and related assets for iteration two, custom asset creation and spot UI reviews run in parallel in support of iteration one.

This  completes the lather, rinse, repeat cycle.

Meanwhile, persona assets, user stories and related assets are aggregated up for incremental release review.  Any usability or user experience hurdles are triaged and assessed for re-insertion into the iteration plan.  Instructive text, user help documentation are written and evaluated for release.

Benefits and Lessons Learned

As a seasoned UX practitioner, I know the value of getting the requirements right before writing a line of code and my initial reaction to agile development was harsh to say the least.  It’s just a temporary jolt.  Once you get in the swing of rapid iteration and continuous design, you barely miss lengthy requirements gathering and documentation.  The clear benefit is low upfront project spend and near term return on investment.  In traditional models, upfront costs on analysis, strategy, definition and design don’t immediately translate to rapid deployment.  And the upfront cost is significantly higher to account for  end to end specification prior to development.

In an agile team, the analysis, strategy and definition are more light weight and design an development run in parallel.  If high yielding business benefits are addressed in the early release, you will be seeing a return on the investment   earlier than you would have if you staggered the design and development in waterfall fashion.

One major lesson learned for Agile UX in practice:  It is absolutely critical to get one or two iterations ahead of the development team. One slip, and you lose any runway for giving yourself the time  necessary to construct successful solution to meet the needs of your users.

Happy Sprinting UXers

-Jon Fukuda

This entry was posted on Thursday, December 17th, 2009 at 11:14 pm and is filed under Agile Usability, Hot Topics, Limina Perspective. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.