Agile software development has become the mainstream for a few years, now. With the help of its successes in smaller and younger companies, larger and more traditional companies like banks are starting to experiment with them as well. Lately, scaled agile seems to be a hot topic. More structured frameworks like SAFe and DAD are becoming more popular by the day. Yet, there is strong resistance, and sometimes almost resentment, towards them from some agile practitioners. Everyone agrees agile needs to work at scale, the disagreement is whether that needs to be done under a seemingly more waterfall-like framework.
Interestingly, both frameworks mentioned above are started with people used to work in a large enterprise environment. Should we trust frameworks that came out of such traditionally non-agile settings? We all know that one of the major reputation issues with agile is people practicing wagile (waterfall agile) or scrumbut. What makes these self made frameworks better than the rest that people look down upon? Well, it turns out, a lot of the original Agile Manifesto signers worked in large enterprises. That proves good things can come out of these environments. In addition, having consulted in one of these enterprises, I can say that people who work in large enterprises are just as talented, creative, and sometimes free spirited as the rest. However, they also constrained by a lot more external factors like regulations, industry rules and standards, social pressure.
An experience agile coach Joe Little had an interesting correspondence with James O Coplien through one of his blog post on this very subject. While James is obviously very knowledgeable in agile and especially Scrum, I agree with Joe’s point that explaining, much less convincing, most people about expanding the agile practices in a self-similar pattern is a quick way to send people to the clouds. It will be even worse talking to business executives like that who already have a preconceived notion that structure and rigid processes provide security and quality.
The biggest advantage I see in these frameworks is that it provides a more gradual transition path from waterfall to agile development. By implementing a more structures top layer (at portfolio or product line or business line level), they provides predictability and discipline required of public companies from the investor. That also gives more freedom to teams and business units to make decisions on lower level details.
It’s a balancing act. Just like one of my favorite quotes in agile, “take one bite at a time”. Doesn’t it make sense to approach agile at a large scale in the same way for some environments?
In a follow up discussion on this post, a number of very interesting points were posted here. Please continue reading the comments. The conclusion/ take away took a slight turn. Instead of evaluating the scaled agile frameworks, it is the detailed things we do and the ways we think that makes a difference. The frameworks, like a lot of other large frameworks, can be used as sounding board to reflect on our ways.