- Aug 24, 2011
I've been thinking for some time about how to leverage lean principles from a UX /product development perspective. I'm finding some facets of UX/Prod Dev work especially those around early stage discovery/research/analysis aren't lending themselves to standard agile methods and tools. I'm not talking BUFD, I'm talking understanding the overall context. I'm thinking the phrase "Lean Kickoff" may encapsulate the concept. I've tried to incorporate things like Dennis Steven's Performance-Capability Model and other QuickStart methods.
How do you quickly get to a well-prioritized backlog for an ill-defined project like "Fix the web site?" In this example, I'm a UX contractor, but am acting as UX architect, consultant, agile coach/champion, and co-lead of the product owner team. I'm trying to look at it holistically: web analytics and metrics; content management; data model; IA/IXD; patterns & components; SEO; accessibility.
Discovery stories are more about learning and less about doing. My current project is a website restructuring/redesign. Instead of stories like "As a shopper, I want to add an item to my cart" I have something like "I want to get visibility into what are the biggest problems on the site" or "I want to know what is on the site" or "I want to know who uses the site most" or "I want to understand what users can do on the site." Tasks end up things like "meet with business unit X to understand their goals and what they do" or "review the existing style guide." How do you define done in a learning context?
Even understanding general business goals like "I want to know what are the goals of the site" or "create a prioritized list of users" gets problematic very quickly (multiple stakeholders). Are lean/agile methods and tools (stories, backlog, taskboard) even relevant in this context?
For example, taking the story/backlog approach, "I want to get visibility into what are the biggest problems on the site" is more of an epic, so I break it down to "get an inventory of what pages are on the site" and "get hit counts for those pages." These look doable, at least till I start digging into them, and discover a whole host of new issues. Like 90% of the pages aren't HTML (most are pdf). We have Google Analytics, but the tracking has only been implemented on HTML pages that have been developed over the last 4 years, which is about 1% of the site. The more you dig, the more questions you get, and most of the new stories are for infrastructure. Adding GA tracking to a group of pages isn't a traditional user story, but it's the type of work that needs to be done first in order to understand what needs to be done next.
Does anyone else deal with these type of issues? What kind of solutions have you found?
- Next post in topic >>