Search This Blog


13 May 2013

Help! They're running an Agile project, how do I know when to worry?

The parents of teenagers and new-to agile project managers seem to have rather a lot in common. There are a few similarities that spring to mind immediately, such as; grey hair and  large mobile phone bills. They are also likely to be experiencing an underlying sense of uneasiness, coupled with cautious optimism and occasionally sheer panic. I have teenagers and am working with a new-to-agile project manager, so I feel qualified to comment.

Neither teenager(s) or Agile projects do well under helicopter style management, but the parent, or PM's urge to operate this way is very hard to resist. It's tough when you find that your customary approach to feeling in control of a situation no longer works and it's very hard to resist worrying all the time, about whether you should be worried or not, especially when the subject of your worry is adamant that its "all good".

In a plan-driven software development project, the number of requirements completed versus the planned requirements completed over time appears to provide meaningful information about a projects status, any deviation from the plan is seen as a red flag. When there are a lot of these, you know exactly when to worry.  In the context of an Agile project, things are not so clear cut, the team may be delivering done requirements on a regular basis, but the extent to which these represent business value is the key indicator. There are obviously challenges in quantifying something as subjective as business value, but this represents the only the tip of this particular iceberg. Both Agile approaches and plan driven approaches have the same goal, which is to improve the outcomes of the software development process but they approach this goal from fundamentally different philosophical perspectives. This philosophical divide becomes very evident when an organisation tries to introduce Agile methods at the team level, but fails to do so at the project governance level.  

When I needed to select a research topic as part of my ongoing studies, I decided that it would be interesting and possibly useful to investigate this. I was particularly interested in finding out how this issue is solved by organisations who's governance and reporting infrastructure is designed to use data from plan driven projects, but which develop software using Agile approaches. This research will take several months to complete and as always step one is a search and review of the academic literature in order to understand current thinking and knowledge.

Whilst there is a lot of fantastic and relevant material in the professional literature (everything which doesn't count for the purposes of ERA) my review includes only the academic literature which presents empirical research and has been published by ERA ranked journals and conferences. Doing this review was an interesting exercise I learned a great deal but the output does not constitute light reading, so for fun, I've winnowed  12 years of literature and the 6000 words I wrote down into 3 paragraphs.

I found that between 2002 and 2006, a lot of research was conducted which looked at Agile transitions in general, while there was some discussion of project governance and reporting most focused on the implementation of Agile methods and the experiences of those at the team level. Where governance and reporting was covered it was identified as an emergent problem. Between 2006 and 2008, a number of case studies were published which did focus on this issue, my personal favourite being a detailed study of the Agile transformation undertaken by an American energy company DTE Energy where the introduction of Agile methods and the focus on delivering business value at the project level is credited with being a key factor in the organisations achievement of CMMI accreditation.

Post 2008 the topic seems to vanish from the academic literature, there are a few papers which discuss very specific ideas such as using earned value analysis (EVA) to report on Agile projects. There are also a number of studies which report on organisations where a move to an Agile approach was  undertaken at the whole enterprise level and where working with stage-gate governance approaches is simply not a factor. In some cases Agile methodologies were seen as the only way to manage the complexity and variability of a specific project, in other cases, such as at Cisco Systems, the use of Agile approaches as best practice was taken so seriously that the organisation invested in an Agile office as an adjunct to the Project Management Office (PMO) in order to ensure that everyone was on the same page. I suspect organisations which take this kind of approach are unusual and that 3 or 4 years ago it was still the case that the adoption of Agile methods was most often driven by software development practitioners rather than from the C suite.

Whilst the academic literature didn't provide any clear answers, of the 20 - 30 papers I reviewed a number of strong consistencies did emerge and included the following, all of which, I think most practitioners would consider to be in the "duh" category

  • Stage-gate approaches to project governance are potentially very compatible with Agile methods.
  • The key success factor for organisations considering Agile implementations is the incorporation of Agile values across the enterprise including at the leadership level.
  • The key failure factor for organisations considering Agile implementation is an inability or unwillingness to engage with Agile values at the leadership level. 
  • All Agile adoptions, are in fact Agile adaptations there is no "one right way" and that a pattern based approach rather than a prescriptive rule set is needed.
  • That Agile coaching is almost more necessary at the C level than at the software team level, since software teams don't seem to have any difficulties seeing the benefits.
There you go, validation if you needed it!

From a practitioner perspective things seem to have changed a bit over the last year or so; entering Agile as a keyword in a job search returns as many results for project manager roles and business analyst roles as it does developer roles.  A quick search of the professional literature suggests that within the project and program management community reporting on your Agile project and the challenges this presents has been a key topic for quite some time now, but is still been driven by practitioner community. This looks a lot like the early majority, perhaps the C'suite will be the late majority, hopefully not the laggards.

At the very least figuring out what is going on here, will very likely keep me out of trouble and distract me from worrying about my teenagers for the rest of the year. 

If anyone would like the full list of refereed papers and articles, please let me know and I will be happy to provide it.