Inital thoughts on Agile with Axure

Part 1: 
The Settings (Or, making software is like making sausage - you don't really want to know how it is made, but you should)
Immediately one faces the fact that there are way more flavors of Agile than Baskin Robbins ice cream...So perhaps we should begin with the key drivers that lead to typical UX projects. Typically it is the business side of the organization that drives the change, with IT group either resisting change, locked into a technology flavor mindset (.Net, Java, etc.) and in many cases the communications between the two part of the organizations are poor.There are many exceptions, but I want to address realities of large scale projects. In established organizations, despite the desire to use Agile, there are still hard wired protocols and bureaucracy around traceability, change management and sign-offs.
Axure, like the typical UX team, sits in the intersection of messy cultures and is being used as a communication tool to articulate, 'sell' and specify both vision and reality - this is where some of the key risks and opportunities are for us as UX practitioners. By that I mean that the initial Axure prototype is probably very aggressive about rich interactions, features and functionality.  The business end gets excited because customers who are exposed to the new vision are pressuring for improvements. The sales department is always hungry for a better product to sell.
Thus, strategic plans and budgets are set around unrealistic delivery targets and often, with very little awareness of the full UX development process. In fact, the UX team is often engaged after the project plan, budget and delivery dates were set (What?! we need to pay for usability tests?!). Moreover, large projects are often broken into phases, which means you can not expect a clean process of creating a prototype, finalizing it and go to the beach, because work on following phases often begins before coding of the previous phases ended - which means that everything is in flux, and the Axure file should be well formed to handle change. 
So who are the players, and consumers of Axure output now the the project is in flight?
  • The UX team (or perhaps you are a single practitioner) - The organizational association of the team is critical - is the team part of the business unit, or part of the IT organization. I've seen both, and, I've seen the team floating with no clear association, which can be worse.
  • The business owners - and depending on the organization, they may be spread all over the US or worldwide, often with conflicting motives and requirements. - The UX team must make sure that UX requirements are captured in a very formal way - Axure can be used for that, but there are some issues around managing the requirements in Axure. While Axure is not a requirement gathering tool it makes so much more sense to capture UX requirements in it that we may get tempted, and have to live with the consequences later on. Not a big deal if we planned for it, a mess if we did not.
  • The IT team - In large organization this may be a Hidra, with sub units that are in charge of some aspect of the technology, they may not like each other, or for that matter, even communicate much on a regular basis. And off shore teams are the norm these days, so we have to keep in mind that the output of our Axure project may be consumed by people for whom English in not a native language.
  • The BAs - In some organizations each group might have their team of BAs - so the documents that are produced include business requirements and specifications, technical requirements and specifications, etc. The UX team typically adds UX requirements and specifications -- All mentioned are often extremely long documents that none really reads because everyone is busy trying to beat the unrealistic deadlines mentioned above. So an opportunity that Axure affords, is generating specifications that are easy to consume, which s is not so trivial, and I'm looking forward for the upcoming resale for some enhancements. Agile does not mean that specs and requirements are not needed - in fact the issue becomes worse - how to track and mange changes from scrums and sprints such that developers and business partners don't get lost in the sea of documentation?
  • The QA team - this team (if it exists beyond being a place holder) has the worst job - testing scripts are often rendered useless because changes happen all the time - so they need to be able to follow up the rapid changes, and update the testing library in time for testing.
This post was originally posted on the Axure Discussion board.