I’ve been asked to offer a public lecture for Ukraine by the DTEK Academy, the learning and development arm of one of the country’s leading energy providers. As we bring 2022 to a close, I thought I would share some of the contents of that talk.
Image source: https://www.npr.org/2022/05/26/1100501433/heres-how-much-it-could-cost-to-rebuild-ukraine-and-who-would-pay-for-it
The tyranny of uncertainty
Human beings are truly terrible at dealing with uncertainty. Our forecasts are almost always wrong, particularly for major, expensive projects. The Big Dig in Boston. The California bullet train to nowhere. The Iridium project. We’re overly optimistic. Sometimes we are politically motivated to push misleading information. And sometimes we just take action because we feel rushed and want to “do something.”
While this matters a lot under normal circumstances, it has huge implications for a project as potentially massive as rebuilding Ukraine in a post-war world. By some accounts, it could be the biggest common project the West has undertaken since the Marshall Plan after World War II. It could be a trillion dollar opportunity for those brave enough to take the risks. This is where my concept of being discovery-driven and Bent Flyvbjerg’s research on mega-projects intersects. Both of us recognize that under uncertain conditions, you have both no basis for making accurate projections and every kind of human cognitive bias that can cloud your judgment.
The only way out of the uncertainty trap is through learning. But here’s the thing – massively failed projects, like massively failed ventures – have certain characteristics in common. Big budgets. Huge promises. Big teams. Plans that look really detailed, but which are essentially documentation of fantasy. No breakpoints to stop and ask whether we are headed in the right direction. Leaders personally (and often emotionally) committed to a single path. And often, objectives that are wildly ambitious. So what is an alternative?
Adopting a discovery driven approach
A discovery driven plan connects the current and future path a project is on in a very disciplined way. It also pushes you to demonstrate how you are converting raw assumptions into knowledge as you go. This applies to not-for-profit and social ventures, as well.
The steps are pretty simple.
- Define what a good outcome would look like, and what your measure of impact will be. For a for profit business, this is usually profits, and the measure of impact is the thing you are selling – your unit of business. For a massive program like rebuilding Ukraine, you would have one fairly big overall goal (something like ‘prepare the country for legitimate membership in the European Union’) but each project would probably have sub-goals. A goal around education might be how many students are moving along a curriculum for skill-building. A goal around energy might be how sustainable and independent Ukraine can become. The point is to start with the end in mind.
- Now, work backwards. What are the logical implications of how many units of impact you will have to create to have the overall plan succeed?
- Still in working backward mode, consider the operational requirements of delivering that many units of impact – I call that a deliverables specification. Document them so that everyone on the team understands what we’re trying to accomplish here.
- As you are doing this, you’ll be making many, many assumptions. Make sure you document these. By explicitly calling them out as assumptions, you empower your team to ask the “why” questions. Why do we believe that lapsed time will only be so long? Can we find evidence that would offer support for our assumptions? Even better, can we run an experiment to find out which assumptions are more correct than others?
- Finally, when planning for a new venture, I talk about planning through a series of key checkpoints. A checkpoint is simply a moment of time in which you are going to learn something useful, in other words, to validate your assumptions.
This is where Bent’s work on mega projects has so much resonance for me, as he advocates for breaking big, customized, monolithic projects down into modular pieces to facilitate learning as you go. It also often means that you are delivering a result far more quickly than if you had to wait for the entire project to be completed before any value could be seen.
Lessons from the Madrid Metro
Bent uses the example of the Madrid Metro to illustrate these points. The systems’ planners used three rules to guide their decision-making.
The first rule was “no monuments.” In many public investments, customized and grand features differentiate one part of a system from another. The problem with that is that you are firstly maximizing uncertainty but secondly have no way of later parts of the project benefitting from what you learned in earlier parts. If you use standardized plans, everything you learn from one part of the system benefits the rest of it, creating the possibility for learning loops and the efficiency that comes from replication. So in the Madrid line, every station was built according to a common standard plan.
Rule #2 was “No new technology.” We all know planners who want to use the latest, state of the art, tech in whatever it is they’re doing. The trouble is that new things are unpredictable and can fail in unpredictable ways. They are vulnerable to the unripe ecosystem problem. Instead, identify technologies that are tried-and-true and innovate, perhaps, in how you combine them.
Rule #3 was “speed.” In entrepreneurship, we are used to moving fast because every days’ delay is a risk of losing money and of competition imitating what you are doing. In the Madrid case, they realized that the rate-limiting step for making progress was how many boring machines could be deployed at one time. Instead of using one or two machines, as was prevailing practice, they at one point at as many as six machines operating in parallel. Now you can have the crews learning from each other, can see substantial progress toward the goal and more importantly, can shorten the total time consumed which has huge benefits for cost and risk.
What does all this mean for Ukraine?
It means, as Flyvjberg points out, spending a lot of upfront time thinking, before taking action. As he argues in his forthcoming book How Big Things Get Done, “Think Slow, Act Fast” should be our mantra. That means spending the “slow” thinking time to really understand the systems being created and how they interconnect. Better to spend the time on that than rush into infrastructure rebuilding willy nilly only to discover problems later on. It’s cheap to fix problems on the drawing board. It’s expensive to fix them in real life.
Specifically, Bent offers a set of heuristics, or rules of thumb, that he suggests can be helpful for those contemplating big, ambitious building projects. I adapted some of these for my talk for Ukraine:
- Work backwards. Start with your goal, then identify the steps to get there.
- Understand your odds. Most big ambitious projects fail – make the effort to beat the odds!
- Plan slow, act fast. Getting to the action quickly feels right. But it’s wrong.
- Find your Lego blocks. Big is best built from small.
- Align your team. You won’t succeed without an “us”
- Master the unknown unknowns. Use reference point forecasting to dispel the myth that your projects are unique.
- Know that your biggest risk is you.
I’m happy to answer further questions and will share the recording of the presentation when it is available.
You can drop in on a recording of my Fireside Chat with Bent Flyvjberg here.
If you’re interested in diving into plans for specific projects to rebuild, a group of economists have put together this fascinating report outlining possibilities.
Have an important, but uncertain, project on your mind?
If so, drop me a note! We’ve been working on several practices, tools, guides and diagnostics that we believe will be game-changing for those companies that adopt them.