Agile or Waterfall? It Depends.

Agile Project Management Real World Series · Episode 1

Good project management is not about knowing one way to deliver a project. It's about knowing which way is right for this project.

The right approach depends on the project

I believe project managers should be agnostic about delivery approaches.

• Don't choose Agile because everyone is doing it.
• Don't choose predictive because that's what you've always done.
• Choose based on the nature of the work, the level of uncertainty and what gives the project the best chance of achieving its outcome.

That sounds obvious.

But in practice, organisations often have a preferred way of working — and every project gets pushed into the same model.

Predictive project management has its place

Traditional, predictive management is a proven, powerful approach. If we can reasonably predict what, how and when upfront, it can work extremely well.

Think about:

• Building a bridge
• Setting up a factory
• Designing the core hardware of a new phone

Detailed upfront planning makes sense when changes later are difficult or expensive.

But what happens when things aren't clear?

What happens when we apply that same mindset to something where the needs aren't fully clear yet — like developing a new digital product?

The customer need isn't fully understood. The market is changing. Users don't know exactly what they want yet. Technology is evolving. And your team will learn more once people actually see and use the product.

Trying to define everything upfront creates a problem. When you use a fixed plan in an unpredictable space, a predictable pattern can occur:

• Requirements, design and estimates are baselined when we have the least information.
• Customers are heavily involved at the beginning, but may see little of the actual product until late in the project.
• Months into the project, we may have plenty of documentation and status reports — but little working product.
• Testing and validation happen late, when defects or missing requirements can lead to significant rework.
• The business changes, customer needs change, or new information emerges — but adapting the original plan becomes increasingly difficult.

Enter the adaptive, Agile approach

Instead of betting everything on one big reveal at the end:

Build → Show → Get feedback → Learn → Refine → Build again

We work in shorter cycles and deliver small, working increments early and often.

Agile doesn't say "don't plan." It says plan continuously as you learn.

With approaches such as Scrum and APMG AgilePM, teams can use iterative cycles to inspect the product, gather feedback and adapt. Plans, estimates and priorities evolve based on real evidence rather than early assumptions. And testing happens continuously, so problems can be discovered earlier.

But Agile isn't automatically better

Predictive isn't automatically better either. Sometimes the answer is both.

A simple example: the smartphone

The hardware needs detailed upfront engineering because changing the physical design later can be extremely difficult.

But the software and operating system keep evolving. A security vulnerability might need an urgent fix. A new feature might need to be released quickly. Customer feedback may change priorities.

Why force the whole project into one approach?

A hybrid approach may be the better fit. This is increasingly relevant in complex environments such as automotive, manufacturing, energy and telecommunications, where different parts of the same project may have very different levels of uncertainty.

The bottom line

So what should a good project manager do?

They ask: "What does this specific work require?"

• Understand the work.
• Understand the uncertainty.
• Understand the constraints.

And choose the approach — predictive, adaptive or hybrid — that fits.

The principle

Good project management is not about knowing one way to deliver a project. It's about knowing which way is right for this project.

How this connects to AgilePM

That's the thinking behind our AgilePM course — not simply learning another methodology, but understanding how Agile projects can be planned, managed and governed effectively.

Explore AgilePM Foundation →View upcoming courses →

What approach does your organisation default to?

Do you mix frameworks, or stick strictly to one? We welcome the discussion — share your experience with your team or with us.

Next episode

Predictive, Adaptive or Hybrid — How Do You Decide?

I'll share a simple way to think about choosing the right approach. Episode 2 will be published here soon.