Wednesday, May 8, 2013

A Tale about a Failed Project: The Death Start II program


When introducing agile thinking I usually show the Norstrom Inovation Lab video as best agile example. Though, to make some of the points come across I also show a typical example of a failed traditional project. I have long searched for a short video illustrating a failed project and accidentally found this gem when watching the whole Star Wars with my 7 years old in the right order. It is the opening sequence of "The  Return of the Jedi" and it is packed with management 1.0 goodness.

Unfortunately, for copyright reasons I cannot upload the video clip for your viewing. Grab a "Return of the Jedi" DVD (you surely have it somewhere) and look at the opening scene.

The Death Start II program


An Unexpected Visit

Darth Vader arrives on the construction site of second death star. He is greeted by an obviously nervous program manager:

Lord Vader "the emperor is not as forgiving as I am" speech (click for larger)
Program manager: "Lord Vader, this is an unexpected pleasure! We are honored by your presence."
Vader: "You may dispense with the pleasantries commander. I am here to put you back on schedule!"
PM: "I assure you Lord Vader, my men are working as fast as they can."
Vader: "Perhaps I can find new ways to motivate them."
PM: "I tell you, this station will be operational as planned."
Vader: "The emperor does not share your optimistic appraisal of the situation."
PM: "But he asks the impossible! I need more men."
Vader: "Then perhaps you can tell him when he arrives."
PM: "The emperor is coming here?"
Vader: "That is correct, commander... And he is most displeased with your apparent lack of progress."
PM: "We shall double our efforts!"
Vader: "I hope so commander, for your sake. The emperor is not as forgiving as I am."
TM & © LUCASFILM LTD. ALL RIGHTS RESERVED. 

Management by Fear

This is about a senior executive sent by a concerned top management to put a program back on schedule. We understand ("back on schedule", "optimistic appraisal", "lack of progress") that top management has lost confidence in the program and decides to take the matter in their own hands.


The program manager is at first trying to defend his position but quickly gives up when it becomes apparent that his program will get undivided top (dark-force) management attention.What we are seeing is a type of management that I would call management by fear. As the story goes, it will be apparent that this is not the best way to manage a project if you want to succeed.

Background

"The Empire" is in fierce competition with "The Rebel Alliance". "The Empire"'s previous flagship product - the death start - was a total failure and got blow to pieces by the competition. Now "The Empire" is desperate for a success. Top management thinks that the death start product actually was good but that the market was not ready for it by then, or that its release was sloppy. Therefore, management decides to launch a new version of the same product - the death start II. Quoting a top manager at "The Empire": "We've done it before, so what could possibly go wrong?".

The Death Start II

Looking at it, we can feel how huge this program really is. It is a massive piece of hardware, looking hugely complex. We see a lot of manpower with heavy security, protocols and processes. We can only glimpse at the nightmarish complexity of the software systems needed to run the whole thing.

Complex? Complex!
The investment cost must be staggering. Actually, thanks to the White House, we have an estimate of the cost of the building a death star: $850,000,000,000,000,000 (850 gazillions?)! If we also include the typical cost overrun of 43% observed by IT projects we end up with a cost of about $1,215,000,000,000,000,000. Well, this is a major investment. It has to be business critical!

Premises

Let us speculate for a moment on the premises on which the death star II program got started:

"We've done it before!". Yes, though the first product was a total failure when launched. It ended quite abruptly and tragically, with many members of the original program disappearing with it (the chief architect/engineer/product-owner, architects, designers, programmers, engineering specialists, independent contractors, etc.). It must have been impossible for "The Empire" to re-form the original team. Therefore, the company had to rely on the documentation left over by the original team, which - again unfortunately - was quite incomplete as documentation was left as a part of the "hardening" activities which were ongoing when the first death star tragically disappeared. Fortunately, the original death star plans were saved by the pre-study project, which was dutifully documented (it is worth noted that the competition has had access to the same plans, though without the financial muscle needed to back-up its construction).

"Last minute change". The program was underway, already somewhat delayed by the lack of knowledge (which the plans approved by top management took for granted), when marketing had a great idea!
It goes like this:
Smart Marketing guy: "We have this great idea! You're going to looove it! Listen: we need to beat the competition once and for all..."
Program manager: "Yes, I am aware of that. That's why make such an investment in this product."
Marketing: "Yeah, yeah! Our idea is that we will soon have this great trade-show when we invite the competition to have a look at our unfinished product. OK? But then ... our product is actually finished! And we blow them to pieces!"
PM: "What are you saying? We show an unfinished product that is... finished?"
Marketing: "Yeah! the thing is that it has to look unfinished, but then it's fully operational! See the genius of it? This is awesome! The competition comes believing that we are not done, but then PANG we're done!".
PM: "Do I get this right? You want us to re-design the product to be functional even when not fully built, to re-plan all projects and sub-projects and tasks AND to deliver earlier?"
Marketing: "Yeah! Come-on! How hard can it be to deliver earlier if you do not build everything?"
PM: *long anxious sight*

So, the program was already delayed due to a lack of knowledge, but now it is even more delayed because marketing sees a business opportunity in releasing a modified version of the product earlier. The fact that it requires an extensive re-design doesn't seem to bother management (or management is not even made aware of it). In this context, we understand that top management eventually has to act by sending a senior executive to take charge of the situation.

The Darth Vader Effect

Let us speculate about what happens after Lord Vader's "the emperor is not as forgiving as I am" little speech. We do not get to see it in the film, but it is quite easy to guess.
Welcome to "Get the job done!" death march:

  • The hugely complex program is suddenly not run as a complex program, but as a simple/complicated one (see the Cynefin framework). Just add more men! Faster! There is no understanding from management that the magnitude of the complexity must be managed by probing and sensing, not by analyzing and categorizing. Vader simplifies it by decree: just to it! This will have a huge impact on the outcome.
  • Overtime for everyone! Stress, fatigue and exhaustion. People become less creative, less willing to cooperate, less prone to see and detect defects.
  • Multiple projects compete for the same "go-to-guys"/experts to complete their tasks in time. Which means more task-switching for the most valuable persons; which means that the most valuable persons actually produce far less. Which inevitably stresses those valuable persons even more...
  • Everything becomes urgent! All plans become obsolete as all tasks are prio 1. This creates conflicts between projects that block each others, and create even more pressure on those few experts that can get things done.
  • "I need more men": forced to show action, the program manager brings in more workforce; which only results in slowing down the productive people by being forced to bring the new ones to speed (Brooke's Law).
  • "I'll fix that later!": Everyone takes short-cuts, building an enormous technical debt! Things are not built to be reliable or with the expected quality. Usually tests the first to be "forgotten", then design, then architecture. The program manager is not even aware of the potentially huge risks that everyone is taking on his behalf.
  • All this stress, exhaustion, task-switching, unprepared work-force and short-cuts are creating a huge number errors, defects and re-work. Quality becomes abysmal.

The Result.


The death star II was delivered on-time!
The program was a success!

Was it a business success?
No! It was a disaster! The product fared even worse than the first time. Speak about a tough market! The weight of the lost investment and the tragic disappearance of some key top managers with the product sent "The Empire" into disarray and eventually irrelevance.

Hitting the market hard...
The first death start became a total failure due to a design mistake (and a lot of luck): a "venting shaft" that connected directly to the highly explosive core. The second product - the result of management by fear - instead has a highway to the highly unstable core! This is clearly the result of stressed, exhausted architects task-switched into taking short-cuts to finish the darn thing in time. This also illustrate the inability of the chief product owners/engineers to control risk on the program.

A highway to the explosive core

Why would we spend countless taxpayer dollars on a Death Star with a fundamental flaw that can be exploited by a one-man starship?
- The White House

What were they thinking?

They were not thinking! They didn't have time.

  • They had to take shortcuts.
  • They had to focus on succeeding with the program, not with the business objectives. 
  • They didn't have a way to cope with changes (both at the feature and design/architecture levels).
  • They didn't have a way to minimize task-switching.
  • They did not understand how to manage complexity.
  • They could not manage risks.
They were not agile, nor lean.

How will you manage your next business critical project?

4 comments:

  1. I find you lack of faith disturbing...

    ReplyDelete
  2. This comment has been removed by the author.

    ReplyDelete
  3. The "Last minute change" part was pure genius.

    ReplyDelete
  4. What I struggle with is the physics going on how can you "build" a death star to start with, surely the gravitational pull on it by the neighbouring planet, never mind itself, would make the whole thing unstable. Sadly yet again the Senior Executives thinking that we can bend the fundamental laws of the universe to suit our own ends

    ..But like the analogy

    ReplyDelete