1. News Bite: $165 Million Blockchain Fiasco at Australian Stock Exchange
In 2015, the Australian Stock Exchange announced an ambitious plan to completely replace their settlement and clearing system CHESS with a new, blockchain-based system. 7 years later, the system was only “63 percent complete” with an “uncertain” timeline to completion. The project has now been completely abandoned. The associated write-down is roughly AUD 250 million (USD 165 million).
I’m fascinated and appalled on multiple fronts.
First, is there any example in history of a 9 figure tech project that didn’t end in a total fiasco? No successful tech company has been built this way – by starting with a hundred million dollar development plan. I don’t think that’s an accident.
When you’re planning a project of that scale then almost by definition you have to take a waterfall approach – there are too many downstream dependencies, which is how you end up with such a large cost number. There is simply no 9 figure project without a Gantt chart.
But waterfall doesn’t work. The reason is simple: “Responding to change over following a plan” is a core tenet of Agile principles, which have been proven time and time again to be the successful approach to software development.
“Oh but we are doing Agile development on our $100 million mega-project!” organizations exclaim. They have Scrum Masters and everything! But scratch the surface and you find it’s really Scrumfall. Because, again, there is no 9 figure project without a Gantt chart.
This is the other big thing that bothers me about this. As someone in the procurement tech space, it just makes me mad on behalf of all the procurement professionals I speak with who can’t get budget for sub-$100k investments that will save their organizations millions. Why? Often because it is crowded out by these mega-projects that promise absurd and unachievable ROIs, and have such high-level support that no one feels comfortable raising their hands to say “maybe we shouldn’t do this”.
Saving millions doesn’t look so sexy compared to an investment that is expected to return billions. But the latter almost never pays off, and in addition to the lost money there is the opportunity cost of not having done smaller projects that would have paid off.
Decades of software development deliver clear lessons for success: start small, demonstrate value, iterate, expand. And yet we continue to see these kinds of bloated mega-projects get greenlit. When are organizations going to learn that it just doesn’t work?
2. Cool Tool: Trello
In the spirit of this week’s news bit, let me give a shout out to a better way of facilitating truly agile management. If you aren’t already using kanban boards, you should be – they aren’t just for software development. And Trello is the benchmark for kanban boards.
The free plan includes a lot – it’s very likely you’ll never need more than that, even with a whole team using it. One tip: make each card very small. And when you think you’ve made it small enough, split it up into several even smaller cards. One big card will never get done, but 20 cards breaking down the individual steps will.
We run Vendorful with kanban – not just the software development team but marketing, customer success, internal projects, etc. It’s a simple and powerful way of making big progress in small increments.
3. Words of Wisdom
“You are maximising risk exposure by doing big batches. Taking big decisions up-front, when you know the least…. no-one actually knows what will work best until we try it, see the result and then tweak it.” – Marcus Hammarberg, Agile Coach