The Remaining Six Lean Principles

The remaining six Lean principles — Build Integrity In, Deliver As Early As Possible, Amplify Learning, Decide As Late As Possible, Empower The Team, and Optimize The Whole — round out the picture started by Eliminate Waste in the previous chapter.

Build Integrity In — integrity, here, means the quality of being honest: what the team actually built is real and accurate, and free from defects or flaws in the system. Lean splits it into two kinds. Perceived Integrity is how the customer experiences the system, and it's extremely critical — the customer doesn't know what went on inside, so their view of quality is based entirely on the interface they actually touch; users believe what was delivered is what they needed, and the overall experience of the system matters as much as its correctness: how it's advertised, delivered, deployed, accessed, and how easy, intuitive, and fast it is to use. Conceptual Integrity is the system's internal architecture and quality, and it's mandatory rather than optional — it means the system's separate components work well together as a whole. If making one small change forces changes in several unrelated places, conceptual integrity isn't there. A system with real conceptual integrity reads, wherever you look into it, as part of the same overall design — not a collection of separate pieces loosely tied together — and it works consistently, has a cohesive architecture, and stays maintainable. It's achieved through the same practices covered earlier in this course: refactoring, continuous integration, automated tests, and daily feedback.

Deliver As Early As Possible — the fastest people survive, not the biggest ones. Early delivery means early ROI, early feedback, and early learning and adaptation, so the practice is to ensure shorter delivery cycles and actively identify and eliminate sources of delay rather than tolerating them as background noise. One underappreciated side effect of speed: with speed, decisions can actually be delayed — a fast-moving team doesn't need to lock in every decision up front, because it can afford to wait and still ship on time. The tools that support this are pull systems, queues, and buffers.

Amplify Learning — learning from what the team actually does, improving using feedback, and building on what's learned rather than letting it evaporate. The practical discipline is simple to state and hard to hold to: don't repeat mistakes, and keep communication frequent enough that what one person learns doesn't stay siloed with them.

Decide As Late As Possible — captured by the phrase "the Last Responsible Moment." Make decisions once things are clearer and once the most information is actually available, based on facts rather than gut feeling. The guidance is to delay a decision for as long as is practical, but no longer — and, notably, it's better to make a wrong decision than to make no decision at all, since an explicit wrong call can be corrected while indecision just burns time. This isn't a contradiction of "deliver early" — the two work together: delivering early creates the real-world feedback that makes deciding later actually safer, instead of deciding late out of indecision with no new information to show for the wait.

Empower The Team — respecting the team's own ability to make decisions, on the premise that the team genuinely knows its own work best. This plays out as self-organization, collective ownership, and shared responsibility, and it produces a motivated, happy team precisely because team members feel valued and the organization is visibly willing to invest in them. The conditions that make this real: no micromanagement, a team that's supported and trusted, recognition and rewards that actually land, and a positive work environment — all of which, together, is what produces a motivated, highly productive team, not the other way around.

Optimize The Whole — a system is greater than the sum of its parts, so improvements need to consider the overall system, not just individual sub-parts in isolation. Optimizing in silos is explicitly discouraged: optimizing each individual part on its own terms tends to produce sub-optimization of the whole, even when every individual optimization looked locally correct. The practice requires understanding how each part is actually glued together with the others. Peter Senge's The Fifth Discipline puts the underlying idea well: "We need to cultivate a discipline to see the wholes, a framework for seeing interrelationships rather than things, for seeing patterns of change rather than snapshots."

Study focus: PMI likes testing Deliver Early and Decide Late as if they conflict — they don't. Delivering early is what generates the real feedback that makes deciding at the last responsible moment actually safe, rather than reckless.