Oct 13 2009
Agility GPS: where are we on the map?
Ulrich Freyer-Hirtz from SQS presented a talk titled The Agility GPS today at the Agile Testing Days conference in Berlin. His Agility GPS is a “systematic and reproducible approach to position fixing on agile projects”, effectively an evaluation system that tells us how agile we are. According to Freyer-Hirtz, this approach can be used to evaluate suppliers who claim to be agile, assess your own project and help at retrospectives.
Similar to the real GPS, this system gives you a position on the map without telling whether that is good or bad. The coordinates of the Agility GPS are the twelve principles of the agile testing manifesto. Practices for agile projects, as well as questions for the assessment, are derived from these principles. The team lists practices that they use or want to use and link them to agile manifesto principles. The model requires the team to assign relative percentage of influence of practices to principles (eg “Pair Programming contributes 100% to motivated individuals”). After this model is created, team members give their estimate of how well the practices are implemented on the project. These contributions are added together and averaged out to get a rating of the fulfillment of various principles.
The result gives you an overview of where you are on the map and points out to principles that aren’t supported enough by current active practices of the team. It then gives you a framework to discuss how these can then be improved, and a later measurement against the same model can give you an estimate how much you improved.
Freyer-Hirtz stressed that this doesn’t give you a magic number to compare your team to other teams, rather an overview of where you are and a list of suggestions for improvement. “Every project needs its own agility model”, said Freyer-Hirtz, and the results of the assessment are related to that model, which is why results aren’t comparable to different projects.
Freyer-Hirtz said that this is still a theoretical model, and it will be tested in practice soon. I find the idea interesting as a systematic approach to evaluate a project, but quite questionable in practice. It relies on self-assessment so it very much hangs on good understanding of agile principles by the team. A team that does not have a good understanding of principles and practices will not be able to get to a good model, and without a good reference model the assessment is going to fail. On the other hand, something like this would be most useful to teams that aren’t agile experts. So it suffers from the chicken-and-egg problem.
One of the goals of the assessment, as stated at the start of the presentation, is identifying “agile alibists”, how Freyer-Hirtz called teams and individuals for which agile development is “I do what I want” rather than following the agile manifesto principles. I don’t really see how this can work for that. These people really don’t have an understanding of the practices and principles so they will simply have 100% in all model cells (or maybe that is the telling sign that something is wrong). A good reference model may help here, but without collecting a number of models from teams that know their stuff it is hard to get to a reference model, and the existence of a generic reference model. I’m really afraid about something like this becoming the basis of certification when you reach a 100%. As Freyer-Hirtz said, the result isn’t a magic number that
is comparable to other team assessments.
In the discussion after the presentation, the ideas mostly revolved around how this model would be useful to teach new teams the principles and practices and help them understand why they are good (eg. what are the benefits of pair programming) and point them towards working with defined practices and principles rather than going by their gut-feel.
For a similar assessment, see the Nokia scrum test.
I’ll be covering Agile Testing Days on this blog, monitor the agiletd tag on this blog for that
Image credits: Tod Koldewijn
 |
 |
Hi,
thanks for the summary! Was there also some discussion about how that self assessment relates to what scrum.org (new organisation by Ken Schwaber and Jeff Sutherland) is offering?
Andreas
hi gojko,
thanks a lot for the summary.
I’ ll generate several “responses” on every topic to discuss that in detail.
as customary, i’ve got the answers to your questions after some hours after the session while thinking about them. thank your asking the right questions, which made me thinking
cheers
ulrich
ulrich
Topic Self Assessment:
the introduced approach is a tool like a screw driver is a tool. it’s up to you to build cool chairs, fine desks, comfortable houses or you may even kill someone with it…
we recommend improving the process of the team starting with getting a common understanding, of what is agile and what are the team goals by defining your agilty model. once, you’ve found your model, you are able to evaluate your practices again and again and contnously adapt the model to the actual needs. it may be feasible to let the initial setup be moderated by a third party.
i believe that looking at the agile manifesto with this systematic and reproducable approach may help any agile team to runaway from daily business and find their specific way
does this makes sense?