“Business needs predictability,” we may sometimes hear when Agile meets enterprise reality. While it is true to a certain extent, this statement itself can be dangerous as people that receive this message may consider it an encouragement to their default modus operandi: trying to fix as many variables as possible to warrant the desirable outcomes. This logic is deeply flawed and is a direct result of Reductionist Mindset. Complex Adaptive Systems operate under a completely different ruleset (see the Dynamics of Complex Organizations). In complex adaptive systems, by definition, the exact outcomes cannot be predicted and any attempt to pursue predictability where it does not belong leads to highly suboptimal results:
- Overload and burn-out of people
- Low-value functionality
- Rapid growth of accidental complexity in the codebase
So what is the right perspective on predictability that a Lean-Agile enterprise should adopt?
Obviously, it is fully dictated by their understanding of how to Exploit Variability and Explore Economic Opportunities. The shorthand answer however is pretty simple:
- Specific outcomes must be considered, but should not be irreversibly committed to
- There’s always a fair degree of uncertainty that cannot be further reduced
- Variability is not bad news, but rather good news: there are always multiple scenarios of advancing economic value, some with very high payoff
Obsession with Predictability is tightly connected with Point-Based Thinking.