IT and Software Development projects don't have a good track record of success. There are many sources for this claim, some credible, some not credible.
The Standish Report(s) report statistics about project failure but have been widely debunked for their data-gathering processes. But there is good data and most importantly the Root Cause Analysis of these failures.
The concept of Root Cause Analysis is a critical success factor in any domain. There are many methods for performing a Root Cause Analysis. Some effective some no effective. The book on the left has a link to the introduction of the process.
The traditional Fishbone diagrams, Failure Mode and Effects Analysis and similar approaches have a fundamental flaw.
For an undesirable Effect to appear two states have to exist at the same time
- A condition that allows the effect to appear
- An action is taken to trigger the Effect
If the condition and or the action are removed the effect does not occur. This is the basis of the Apollo Method, used in domains where failure is serious. Nuclear Power, Manned Space Flight, Petrochemical Plants, Healthcare.
The First Issue
When we hear any solution to the IT project failure problem and there is no Root Cause for those failures that can be confirmed to remove the Condition or prevent the Action that produces the Primary Effect, then any suggestions that the corrective or preventive actions have no basis to be credible.
Here's an example of a short RCA for an upcoming conference on Risk hosted by PMI Toronto. I'm one of the speakers and the host of the Symposium is a colleague where we've worked in several domains on OSHA 1910.119 Process Safety Management process. This domain contains processes, software, regulations, and other sources of conditions and actions that create Effects.
The question early in the development of the Symposium - which is focused on Risk and the Management of Risk. The question is Why Do Projects Fail. Here's the RCA for that question. This is a quick and simple result, requiring more work before an actual set of corrective and preventive actions can be applied. But it shows the principles of applying the Apollo Method the Effect.
- The action is there is No Formal Risk Management
- The condition is Requirements are Poorly stated
And this the Root Cause of Projects Failing. It's not that simple of course, but this is a start.
These, of course, are too simple, but this was done to show a process, rather than an actionable result. The primary message here is simple
We cannot propose any solution to any problem without identifying the condition and the action that creates the causes of the Primary Effect
The Second Issue
When we hear something like we have a method that allows to show up on time every time and claims this can be done without addressing any of the root cause that creates the Effect of being late, over budget, and not having the outcome meets the needs of those paying, we have to ask - does that suggestion make sense?
Not likely
The Punch Line
When you hear a conjecture that some process - say #NoEstimates fixes some vague problem, here's a starting point for finding the root cause
