This is part of an ongoing series. It continues from Coming up with a business idea, part 2. Visit Walking, Peanuts, and Systems if you’d like to start at the beginning.
To find a root cause, I want to examine all of the leaf nodes and see which nodes affect the entire tree if changed. Here is the tree I’m considering:

There are potentially multiple nodes that could be a root cause, so the next step is to pick one that seems to be addressable.
Reality
Estimates contain uncertainty
There are always new projects to start
Cutting corners speeds up delivery
These are causes that are almost true by definition. I wouldn’t necessarily have to “fix” them, but choosing any of these as a starting point would be a last resort.
Unlikely to Change
Projects start with estimates
Delivery dates rely on estimates
Future work depends on previous work
People ask for tight windows of uncertainty
We could theoretically change any of these, but that would require drastic changes in how projects normally operate. For example, we could choose to start a project and deliver when it’s finished without committing to any type of delivery date. That seems very unlikely to ever be a normal way of working.
Non Root Causes
We use conversations to resolve uncertainty
It is uncertain what work is important
People want to be seen as reliable
These are not root causes. Even if I fully solved them, other parts of the tree would be unaffected.
Possible Areas to Focus
It is hard to predict how shortcuts affect future work
This doesn’t really make sense as a root cause. If it were truly a root cause, then projects would only go off course because of shortcuts. However, projects regularly go off course for many other different reasons.
This suggests that my CRT is not 100% accurate. This cause is part of a self-reinforcing loop. So rather than being a cause, it is an accelerator. I could change the cause from We are under time pressure to make estimates
to We are under INCREASING time pressure to make estimates
.
Again, this change vibes with what I’ve seen in practice. As we deliver late, we get more pressure to shrink the time on future projects. With less time to estimate, we make more mistakes, the project is delivered even later, and time pressure increases yet again.
It is difficult to capture uncertainty
This seems like a good root cause. It doesn’t feel like I’ll have to work around a fundamental law of nature. If I solve it sufficiently, that would improve or solve all of the original complaints.
My next step is to further evaluate my potential root cause and gain confidence that it is truly a root cause worthy of solving.