Is Delivery an Art or a Science?
We often look at delivery as an art. But are there more rigorous perspectives that could benefit the field?
We talk about delivery like it’s an art.
Not art as in “beautiful,” but art as in: a craft you learn through apprenticeship, intuition, and scar tissue. We trade stories. We compare methodologies. We debate rituals. We hire “strong leaders” and hope they can bend reality through force of will.
And to be clear: that whole universe has value. I’m not here to dunk on books, frameworks, or the people who’ve dedicated their careers to making complex work go better.
But here’s the thing I can’t unsee anymore: Complex delivery doesn’t just feel hard. It behaves according to constraints.
Projects fail in ways that are weirdly predictable in hindsight, even when everyone involved is smart, well-intentioned, and working hard. That predictability is what makes the postmortem such a familiar genre. There’s almost always a moment where someone says some version of, “We should’ve known,” and everyone nods—because they did, in some hazy way. They just didn’t treat what they knew as structurally binding.
That pattern is what pulled me toward a different lens:
What if delivery isn’t primarily a matter of taste, talent, or vibes? What if it behaves more like engineering under constraints?
Not “engineering” in the narrow sense of software or machines. Engineering in the broader sense: working inside a system with limits, tradeoffs, feedback loops, and irreversible costs. The kind of domain where certain moves are reliably expensive, and certain conditions are reliably fragile—even if you’re very good at what you do.
If you’re tempted to roll your eyes at the idea of “laws of delivery,” good. You should. The last thing the world needs is another guy on the internet declaring commandments.
I’m not trying to present a new doctrine. I’m trying to name something else: convergence.
Look across the landscape of serious delivery thinking—things that emerged in different eras, in different industries, for different reasons—and you start to notice a recurring shape.
- The Theory of Constraints tells you to stop pretending the system is unlimited and focus on what actually governs throughput.
- Lean takes aim at waste and insists that flow, not activity, is what matters.
- Kanban makes work visible and turns management into a question of capacity, queues, and WIP limits.
- Agile (at its best, not as a corporate costume) is about navigating uncertainty through iterative learning and adaptive planning.
- OODA centers speed of learning and decision-making in contested, changing environments.
- PDCA treats improvement as an explicit feedback cycle rather than a good intention.
- Cybernetics—the deep end of the pool—asks what it means for a system to sense itself and self-correct over time.
These aren’t identical philosophies. They often come with different language, different communities, and different failure modes when misapplied.
But they rhyme.
They keep circling the same gravitational themes: constraints, flow, feedback, adaptation under uncertainty, the cost of delay, the cost of coordination, the limits of human attention, the dangers of invisible work, the difference between “busy” and “effective,” the way local optimizations can hurt the whole, the way systems create the behavior they later blame on individuals.
When a bunch of independent disciplines keep rediscovering similar dynamics, it usually isn’t because everyone is copying each other. It’s because they’re bumping into something real.
That’s the orientation shift I’m interested in:
Maybe delivery has a deeper structure than our current conversations tends to acknowledge.
Maybe the role of methodologies isn’t to win arguments—it’s to offer partial tools for dealing with recurring constraints.
In other words: I don’t think the books need to go out the window. I think they’re pointing at something that hasn’t been unified yet.
This matters because the default alternative—treating delivery as primarily a leadership problem—creates a predictable trap. If success is about leadership and effort, then failure is about insufficient leadership and effort. We respond by hiring new leaders, reorganizing teams, escalating pressure, and demanding more optimism.
Sometimes that works for a while. More often, it just pushes the system harder against its constraints until something gives.
In constrained domains, willpower is a poor substitute for design.
Good delivery, in my experience, is less about heroic pushing and more about making success mechanically plausible. It’s what happens when we respect the shape of reality early—before the plan hardens, before the budget becomes emotional, before momentum becomes identity.
And if that’s true, it suggests a missing piece in how many organizations talk about delivery: we infer confidence instead of assessing it.
We approve initiatives because they are desirable, because the business case “makes sense,” because the political moment is right, because the team seems capable, because everyone is tired of debating and wants to move.
Then we act surprised when the delivery system behaves like a delivery system: limited capacity, delayed feedback, compounding coordination costs, and irreversible decisions made under uncertainty.
A more mature posture might be: Before we move, can we name what must be true for success to be plausible? Not as a slogan. As a real assessment. As an engineering question.
That’s where my curiosity is heading.
I don’t have a neat set of “laws” to hand you today, and I don’t think I should pretend I do. If there are universal principles worth naming, they should earn their way into language—through testing, through falsifiability, through surviving contact with messy reality.
But I’m increasingly convinced of this much:
When delivery disciplines converge, it’s pointing to constraints. And constraints are the beginning of wisdom.
If we can get clearer about the underlying constraints that govern delivery outcomes—across industries, methods, and personalities—we may be able to do something that’s currently rarer than it should be:
Build delivery confidence before the world demands a rescue.
Stay tuned.