Back in 2004, when he was a vice presidential candidate, then-Senator John Edwards hit a nerve with his “Two Americas” speech at the Democratic National Convention. Here is the core sentence in that speech:
And we have much work to do, because the truth is, we still live in a country where there are two different Americas, one for all of those people who have lived the American dream and don’t have to worry, and another for most Americans, everybody else who struggle to make ends meet every single day. It doesn’t have to be that way.
The Agile movement faces its own version of the Two Americas problem. It’s appropriate and necessary to celebrate Agile’s success. There’s no better sign of it than this year’s Agile 2014 conference, which featured both record attendance and a rich agenda. The wealth of good conversations in hallways between sessions, on the bus from the overflow hotel to the conference, and the chance encounter over breakfast or a beer were other good indicators.
Under the right circumstances, wealth can spread through the population naturally. Under other circumstances, it can just as easily concentrate in the hands of a small elite. For the Agile community, the question in 2014 is, What deliberate steps must we take to spread the benefits of Agile to new places, beyond where it already is?
The rapid development and adoption, in some organizations where Agile has already taken root, of adjacent approaches like DevOps, are profoundly valuable developments in those settings. It should not be surprising, therefore, how much DevOps content there was at the Agile 2014 conference. There’s a definite hunger for it among people who want to move from adopting team-level Agile practices to making complementary improvements in other parts of the software value stream. The continued popularity of scaled Agile presentations is an equally healthy and natural development. People in a successful Agile team want to see what it will take to capture that lightning in a bottle and take it, not just to other groups, but to the scale of the larger software development and delivery organization.
Both DevOps and scaled Agile, however, are what in modern parlance you might call “First World problems.” They represent the challenges facing people who are already successful, and comfortable enough in that success to ponder what to do next. What about the places where Agile has not taken root yet, or has found the soil to be far less hospitable than, say, a San Francisco-based Internet start-up, or the interactive marketing department of a retail company?
We’ve seen some discussion about how to plant Agile in less arable terrain. Often, it has appeared under the topic description, “Agile in highly regulated environments.” While that’s a step forward, it’s a highly tactical way to look at a much more strategic problem. You can devise approaches or adopt tools that will help deal with 16-ton documentation requirements and satisfy auditors. These low-level innovations cannot deal with scenarios like the following:
- A large manufacturing company with a highly siloed IT organization that is struggling to move away from a heavy-weight software development lifecycle (SDLC).
- A large software company, in the market for decades, buried in a crushing load of technical debt.
- A government agency that has outsourced some portions, or every part, of every new development project, to satisfy externally-mandated budget restrictions.
- A financial services company with a heavy reliance on a mainframe back end for its system of record, and a poisonous level of distrust between the business and IT.
For Agilists, are these scenarios worth pursuing? Or are they not worth the effort?
I’ve already spoken on this topic, both at last year’s Agile conference, and during part of this year’s analyst panel discussion, in which I was privileged to participate. You can hear the longer version of my argument from those venues. This blog post is just long enough to pose the question, which is really more of a challenge.