One of the tough challenges of acting as a strategic software architect is that it’s not precisely an understood job. Many people ask me what I do, and after several fumbling minutes, I point them to this blog.
The most recent analogy that became somewhat useful is that strategic software architecture is about playing chess while the rest of the world plays checkers. And what I mean is that the world is looking at short time horizons, planning the next step, while the role requires planning two or more steps—and simultaneously making moves during the current phase.
Strategic software architecture is not about the stuff that is shipping now. It’s about the thing that will ship in the future. And the job is not to deliver the current thing. The job is to make sure that the team can deliver the current stuff without you.
As that strategic software architect, when you are working on an immediate deliverable, what it means is you failed your team a long, long time ago. If they need your help, it means that you either didn’t provide them the resources, the people, or strategy that would ensure their success. Fail enough times, and there is a new strategic software architect.
In most software companies, the planning cycle doesn’t extend out longer than 18 months. The world changes too fast for anything more than that. And so nobody is thinking past 18 months.
There is one group that is thinking past 18 months, these strategic thinkers. They are a critical, sufficiently unrecognized group across a large number of business functions, but this is about engineering, so I am focusing on that. As this role doesn’t exist and isn’t recognized, and there are no rewards for long term success and, it begs the question of ‘does it exist’?
There are many people like that at a company. They are the ones who seem to generate magic just when the company needs it. That continues to deliver value, and nobody knows why or how. As engineers, they do it with well-timed code-reviews, speaking whispers to the right people, working on the right project, checking in something that nobody expected. They call meetings to discuss things in private, and thereby create a social network that is impenetrable and built around the respect they have earned and the reputation they have acquired.
And over time, the company strategy is the strategic thinkers’ strategy, even if the company thinks otherwise. For many reasons, beginning with hiring that is shaped through their biases. What is easy to build and hard to make is what they and their social network think is easy and hard. What can be created is controlled by their tastes. What is easy and hard to do is shaped through the myriad of small technical decisions that make change very hard. And their software architecture ossifies their decisions through org charts that can endure long past the code choices that formed them are relevant.
Currently, this entire area is left to chance. We are lucky to hire people who can do the job. And I have seen it in my career. Where there are groups that seemingly out of nowhere, keep doing the right thing. And things keep getting better, but I can’t figure out why. And finally, somewhere someone turns up that has a plan . That plan exists in their heads, or on a piece of paper or a confluence page that nobody reads but that everybody is working on.
I believe that a company that figures out how to do strategic software architecture as a discipline and incorporates it into the 18-month planning has a decisive advantage. I also believe that if they can couple this approach with a rabid focus on immediate delivery, they can’t lose.
Leave a Reply