Agile to Waterfall 2. Waterfall Bites!

Episode 75 - 08 Mar 2017

Last time, I told the tale of an Agile process that morphed into a Waterfall process.

(If you missed it, here's a link )

The new Waterfall process was more efficient. No doubt about about.

But there was a problem.

When things went wrong, they went badly wrong.



My process evolved over a period of months from Agile to Waterfall.

The new process was more efficient.

But it wasn't long before I learned a hard lesson.

Waterfall ... can bite.


This is Professor Professor David Myddelton of Cranfield University.

What little I know of accounting and finance I owe to this man.

His classes were borderline terrifying: woe betide anyone who failed to prepare for a lecture.

But Professor Myddelton was not without a sense of humour.

Like the time he said that is was trivial to make a fortune on the stock market.

Really? He had our attention.

He told us the that it was so trivial, that only two things had to be true.

Number 1: you have do to something different from the the rest of the market.

We couldn't think what the second thing would be.

Some of the Investment banking types were brave enough to take a stab at it.

He waved each suggestion away.

Finally he put us out of our misery.

You have do to something different from the the rest of the market.

And, you have to be....

You have to be right!

Hmmm. Great. Thanks professor.

Agile to Waterfall

In the previous episode, I described how the process I use for making videos has changed over time:

  • Man makes videos in an iterative fashion. Let's call it an Agile Process
  • Man is unhappy: each video takes 10-20 hours to make
  • Man optimises process
  • Man ends up with something that looks like a Waterfall process.

Mike Jones had a much more eloquent review:

"You have shown us how to apply lean principles by reviewing your end-to-end value stream and removing waste and rework."

Mike, you're absolutely right: my process has come a long way since those early days.

My new Waterfall process works well. Really well.

Except when it doesn't

And when it doesn't...

Well, let's just say that those are the weeks my family still talk about.

Let's change this a bit. Some brighter colours. Make the circles into labels. And change the length of the bars.

Writing the script is the longest battle: it develops over the course of several days.

Recording the audio is quick.

Visuals. And then editing it all together.

Yeah. That looks about right.

On good week, that's more or less how it pans out.

On a bad week, the visuals takes waaaaay longer

I'm not taking it taking 10% more work. I'm talking 100% more work.

I don't have the luxury of taking my sweet time about it.

The videos go out on Wednesday at 3pm come hell or high water.

I've got myself a cast iron, fur-lined, ocean going HARD DEADLINE.

I'm committed. I have to find the time.

Even if it means getting up at 4am.

I'm sure you can guess what this does to my mood. I never need to tell my family when I'm in this situation. They know.

And the worst thing about the whole sorry situation?

It's that I'm acutely aware that this ugly pile didn't happen by accident: there's a clear cause and effect here.

Some bloody idiot wrote something into the script over here...

... that led to all the extra work over here.

I am, quite literally, the author of my own destruction.

(literally the AUTHOR. Not quite literally destruction. Although sleep deprivation makes me feel pretty destroyed.)

Fighting back

It turns out that fixing this is trivial.

So trivial in, in fact, that only two things need to be true:

  1. The script has to be completed in a timely fashion.
  2. And... anyone? anyone? it has to be... It has to be RIGHT!

Right for what?

Well, right in and of itself:

  • a half-decent story
  • with a start, a middle and an end.

That kind of thing.

But also "right" in a less obvious way: right for the next process.

Scriptwriter Gary needs to be aware of the visual implications of what goes into the script.

If he puts something in there that requires a ton of video work, then Designer Gary is going to have a bad week..


ScriptWriter Gary is about as well-positioned as it's possible to be to produce the "right" script for Designer Gary.

Genetically perfect. Born to do it.

And yet, ScriptWriter Gary manages to create a script that is "right" only about 60% of the time.

Not a very impressive hit rate.

You've heard it said - possibly on Development That Pays! - that Waterfall works well for processes that are well-understood.

I guess I hadn't understood just HOW "well-understood" the process needs to be.

Confession time

I'll end today wth a confession.

The video you are watching was NOT produced using this process.

I ditched that process weeks ago.

I've swapped efficiency for something that's much more important to me.

Can you guess what it it?

I'll reveal all... next time!