Software Has Outgrown Project Management (And PMI Knows It)
The same software that runs your morning train, powers your VISA card, and keeps airplanes in the sky has evolved far beyond what traditional project management can handle.
"Captain's Log, Stardate 2024: We're lost, but we're making good time according to the project plan!"
If this sounds absurd, that's because it is. Yet this is exactly how we're running most software organizations today - following plans that become obsolete the moment we write the first line of code, while pretending we're in control because we're "on track" according to some project management tool.
Here's the reality: Software development has fundamentally changed in the last 40 years, but the way we manage it is still stuck in the early 20th century. Project Management - that recent invention from the industrial age - is bankrupting our industry, both literally and figuratively. Need proof? Just ask KMart, whose bankruptcy was directly linked to a failed IT project.
The Map Is Not The Territory
This famous quote from Alfred Korzybski perfectly captures our current predicament. Project plans, Gantt charts, and detailed timelines are just maps - increasingly inaccurate ones - of a territory that's constantly shifting under our feet.
Project plans are - at their core - wishful thinking, as in "we wish reality would unfold this way", but when we start executing any plan, we will face unexpected events.
These unexpected events will challenge our view or reality, but some will cling to the illusion, like USS Enterprise's Captain Janeway momentarily does in "Persistence of Vision" (Star Trek: Voyager).
However, Janeway and the Doctor eventually determine that they were under the spell of a telepathic species.
Just as we are in the software world: under the control of a telepathic species, which prevents many of us from seeing that Project Management (even when useful!) is at its core an illusion.
Software gets delivered by people, and machines working together, not by plans, or even by following a plan carefully. Software happens in reality, not in a plan.
Just like a carpenter may have a plan for a chair, but ultimately must work with the wood they get from their supplier.
We, software people, must work with the reality of delivering software every day, not with the plans made when we knew the least about what we were about to deliver.
The project plan is the "map", but we must navigate the actual "territory" of delivering software, every day!
From Alchemy to Chemistry: Our Industry's Evolution
Just as alchemy gave way to chemistry, our software industry is going through its own (r)evolution. We started with project management - our industry's alchemy - but now we need to embrace empirical methods that actually work for software development.
I remember readying "Agile Software Development with Scrum" by Schwaber and Beedle, and reading the chapter on empirical process control (fittingly coming from Chemistry), and being transformed in my thinking.
There I was, as a project manager, learning about how the Chemical industry used sensing (inspecting), and adapting cycles in the scale of seconds or minutes to produce some of its most advanced produces.
The importance of empirical process control cannot be overstated, even as we may want to benefit from plans (like project management tries to do), we must never "get high on our own supply" and believe those plans.
Empirical process control is fundamentally a pragmatic cycle: do, check, act. Even if we must "plan" first, the last three steps are critical because software is running our world!
Software isn't just another industry project management can "optimize" - it's eating the world:
The train you took this morning runs on software as much as it runs on tracks
Your VISA card is just a key to a distributed database
The plane you flew in last month? It's a computer with wings
We, the Agile practitioners, need to continue to improve how software gets developed, with software-native ideas!
The Great Merger That Changes Nothing
If you're not an active member of the Agile community, you might have missed that PMI and Agile Alliance have merged. This is a massive earthquake for the software world, because software is now one of the largest industries in the world, and it's growing fast and project management is not the best technology we have to manage that software delivery.
Agile is expanding, no wonder PMI merged/bought/partnered with Agile Alliance. Agile is now the default in one of the biggest industries of our civilization so far: software. And PMI wants to "own" that space.
While from PMI's perspective this merger makes sense (they now "own" Agile Alliance), it is a threat to the software industry, and the world at large!
The problem isn't that we need better project management for software - it's that we need to stop managing software development like a project altogether.
The Real Alternative To Project Management
Instead of "better" project management, what we need is a mix of:
Methods adapted to software-powered businesses (the largest companies on earth today are software+hardware for software companies)
Continuous delivery/integration/testing, not "better" or "more" estimation and planning
NoEstimates-like approaches with a focus on value instead of arbitrary deadlines
I've been helping teams deliver software without estimates, and without traditional project management, and with much better results. How? By focusing on what matters: delivering working software to users, measuring actual progress through real usage and value delivered, and adapting based on empirical evidence rather than plans.
Time to Take Responsibility
We, the Agile community must take ownership of the challenge that software presents to thew world today.
Software has more impact - for good and bad - than any industry before it. We can't keep using management approaches from the age of steam engines to build the systems that run our modern world.
This is why we're coming together at the Global Agile Summit in Tallinn - once a thriving gateway to the East, now at the heart of a software (r)evolution.
We need to create and document solutions that work for software development, not just adapt industrial-age project management techniques.
The Wake-up Call
Remember: The software industry isn't just another market for project management tools and techniques to conquer. Software development follows different rules, requires different approaches, and needs its own language for management and risk assessment.
Project Management has given us useful ideas, but as a framework for delivering software, it's as bankrupt as KMart.
The future belongs to those who understand that software development is about discovery, learning, and adaptation - not about following outdated maps while pretending we're not lost.
Are you ready to be part of the solution? Join us in Tallinn. The future of software delivery is being written right now, and this time, we're not following someone else's map.
