In our domain there are three common root causes for program difficulties when the program overruns it's budget, shows up late, or the delivered product doesn't work as required
- They couldn't know - the source the problem was in fact unknowable.
- They didn't know - the source of the problem was knowable, but the effort to discover it was done.
- They dodn't want to know - the source of the problem was there, but if it was made visible the project would be canceled or not started
When I read about decision making processes like ...
I'm struck by the 3rd statement. Knowing something about the cost of a decision, the outcome of an investment, the risks, scope impact, progress reporting of future values requires you make estimates. Since all project variables are random variables that interact in random ways - technically non-stationary stochastic processes - knowing about the impacts from deciding anything using these variables means estimating the three core variables of all projects, shown below.
When I hear the suggestion that decisions can be made in the absence of those estimates, think of the Ostrich, ask the following:
- Can I decide about some future outcome without estimating the impact of that outcome?
- If I'm going to invest - spend money - can I do the ROI calculation in the absence of the estimate of the cost or the value - since neither of those are known on day one?
- Risk - by it's very definition - is an uncertainty. These uncertainty are probabilistic outcomes of future events. Estimates are needed.
If there ways to decide these things without estimates, let's hear them. Each of the three variables and each of their drivers is a random variable whose value is not know in the present, but can only be knowing with an estimate of it's future possible values.