Project Mgmt should be viewed like writing a Computer Program

Sometime ago I was starting a big tech project and put all the high level tasks and milestones in a Gantt chart on Clickup.

Then i started adding 2nd and 3rd level tasks underneath those. Many of which were just assumptions at this point. But i put in target completion dates, assigned tasks to folks, etc.

My colleague found it and was like...

"Ken... why the hell are you doing all this? We barely started..."

And I was like.. "Well things will move around but at least now we have a roadmap to work against."

Projects in the company were typically managed with slack and email updates, and not using any project mgmt tool. And typically ran late.

Why did I do this? Let me take you back to 1997...

In 1997 I was a young study abroad student at University of New South Wales (UNSW) in Sydney, Australia. I had taken a programming class and only found out later that the university was quite known for its computer science and attracted programmers from around Asia.

So instead of just surfing and getting drunk each night as I hoped..

...i needed to spend much more time than i had expected sitting in a computer lab (remember those.. haha)

My All Nighter at the Computer Lab

I found myself working almost all night on one class project. It was some computer program, i'm forgetting the specifics. But I just couldn't get it. It had too many levels of complexity.

And I hadn't really structured anything... I'd just dove in and started writing code. But despite the long hours... I'd barely made a dent.

My Computer Programming friend to the rescue

Luckily I had another study abroad friend who was a programmer. So I asked him the next day for his help.

The guy took the problem, and broke it down to its next level... 4 sub-problems.

Then we discussed each sub-problem and broke those each into 2-4 sub-sub problems.

I now had about 12 sub-sub problems that all had a finite scope that seemed pretty easy to code. And so all i needed to do was code the 12 'relatively easy' problems and put them together.

I was floored... this guy had cracked the problem in like 15 minutes flat... a problem that i'd wasted my whole night in the computer lab the day before and had barely made a dent in.

Note that this guy later finished both Ecole Polytechnique in France (the best engineering school in Europe) and had also done a Masters in programming at MIT. So he was also a pretty smart dude in all fairness.

But from that experience i realized.. the best way to solve for complexity is to break it down into sub-components.. until finally you get to a level that is relatively simple to solve for. And then you just solve for all of those parts till you're done.

Or to put it simply.. you use a 'tree structure'.

Which is why... I never try to manage a complex project without some type of structure now. As it is the equivalent to me.. of wasting my night in that computer lab back in 1997...

...when I should have spent getting hammered at Coogee Beach. LOL.