Waterfall vs Agile: VALUE and RISK Revisited

Episode 83 - 03 May 2017

The last two episodes must be the most "commented upon" of all time. It's been difficult to distil the input into a single video.

In fact, it was impossible. We'll need to revisit some of the points raised another time. (It's time we moved on from Waterfall and Agile.)

But I did end up with three graphs that I'm happy with - at least for now. The graphs are:

  • Customer Value
  • Risk of not Delivering
  • Risk of Failure

You'll find Mike Jones' comments here.



A valuable discussion about value, and a chance encounter with risk


A shocking display of shameless bias.

And stick around to the end to grab your free Waterfall vs Agile Cheat Sheet!


This episode, I have to tell you, has been a nightmare to put together.

The number, quality and breadth of the comments on the last two episodes has been amazing; you given me far more than I can do justice to in a single episode.

Mr Sammotube and Mike Jones: you two in particular have, err, challenged me.

Mike put forward a compelling argument about the VersionOne "gang of four" graphs.

He says they're not intended to be "accurate" or "technically correct".

Their strength is their simplicity, and - more importantly - their reasonableness. A reasonableness calculated to "draw in" the old school.

(Here's a link to the previous episode so that you can find Mike's comments. Well worth a read.)

We can't know for sure if there's intentional "bias" in VersonOne's graphs.But it did make me wonder how much "bias" I'd built into my version of the graphs.

Mr Sammotube tested my unconscious biases further. Referring to the Customer Value graph, he argued that the value gap could be reversed.

Really? This is Henrik Knibergs work we're talking about, you know!

Of course it is possible. Given the right project. The right KIND of project:

  • For a complex project, this version stands.
  • But for a more straightforward project - what I tend to refer to as a "Well-understood project" - the waterfall project may well deliver more value at the "big bang launch".


What about Risk?

Here's my "Risk of Failure" curve from last time.

The Agile curve wasn't too controversial.

As for the Waterfall graph, I expected push-back... and I got it. For this part of the line.

There are competing forces at work here:

  • Market changes force the risk up.
  • Any kind of "knowledge gain" forces it down.

Mr Sammotube pointed out that real-world waterfall project teams aren't cut off from the world. There's some level of input, some learning.

So I'm prepared to revise my line down. But not too far. I'd miss the shock value of the upward slope.

This is, of course, the complex project view. For a well-understood project, the initial risk is lower.

Here's the Agile risk profile.

Could it be that the Waterfall risk profile... looks like this? Hmmm.

A different kind of risk

By the way, there was general agreement than the VersionOne risk graph refers to a different kind of risk: the "Risk of Not Delivering".

That being the case, I'm okay with the Waterfall curve, but would prefer an "S" for the Agile curve.

Again, this is the complex project view. I'll leave the well-understood project version as an exercise for the reader :)

I'm also going to side-step a THIRD kind of risk - the risk of missing a launch deadline - chiefly because I've yet to figure that one out.

(Seems I'm also having trouble with split infinitives !)

Case closed?

What started as an upgrade to some graphs on a Cheat Sheet has blossomed into a very interesting discussion .

Case closed? No.

But I do have three graphs that I'm happy with. For now.

Let's roll them in:

  • Risk of Not Delivering
  • Risk of Failure
  • And last but my no means least: Customer Value. With it's all-important "value gap".

That's the complex project view. What about the well-understood project view?

What well-understood project view?

Did I say I wasn't biased?