why methodology is a choice, not a rulebook

A methodology is just an agreed set of rules for how a team moves through the project life cycle described in What Is Project Management?. None of them is universally "better" — each trades flexibility for predictability differently, and the right choice depends on how well the end goal is understood up front.

waterfall: plan fully, then execute

Waterfall finishes each phase before starting the next: all requirements are gathered before design begins, all design is finished before building starts. It produces a very predictable schedule and budget when requirements are genuinely stable — a bridge, a regulatory filing, a physical construction project.
Its weakness is rigidity: if a requirement changes midway through, the fix usually means re-opening an "already closed" phase, which is expensive. A construction firm doing a strip-mall build-out uses Waterfall well because the blueprint rarely changes once concrete is poured.

agile: plan a little, build, learn, repeat

Agile breaks the work into short cycles, each producing something usable, then adjusts the next cycle based on feedback. It trades the certainty of a fixed upfront plan for the ability to change direction cheaply. Agile is covered in depth starting with Agile Fundamentals.
A mobile app team building a new onboarding flow benefits from Agile: user behavior after each release tells the team more than any amount of upfront speculation would.

lean: eliminate anything that does not add value

Lean is less a project-scheduling method and more a philosophy borrowed from manufacturing: identify every step in a process, and remove the ones that do not directly create value for the customer. Applied to projects, it shows up as "why do we need this approval step" or "why are we building a report nobody reads."
A hospital streamlining its patient-intake paperwork is doing Lean thinking — the goal is fewer forms and fewer handoffs, not a faster schedule for its own sake.

six sigma: reduce defects with data

Six Sigma focuses on reducing variation and defects in a repeatable process, using a structured five-step cycle: Define, Measure, Analyze, Improve, Control (DMAIC), covered more in Quality Management. It is most at home in manufacturing and back-office operations where the same process runs thousands of times and even a 1% defect rate is expensive at scale.
A call center trying to cut billing errors from 2% to 0.1% is a Six Sigma problem, not an Agile one — there is no "new feature" to ship, just a process to tighten.

side-by-side

MethodologyBest fitWeak spot
WaterfallStable, well-understood requirementsExpensive to change course midway
AgileUncertain or evolving requirementsNeeds a team comfortable with ambiguity
LeanCutting waste from an existing processNot a full project-scheduling framework by itself
Six SigmaReducing defects in a repeatable processOverkill for one-off or creative work

practical notes

Most real organizations blend methodologies rather than picking one purely — a "waterfall" hardware rollout might use Agile sprints for the software component, or a Lean review might kick off a project that then runs on Scrum. Treat the labels as a vocabulary for describing tradeoffs, not a contract you must follow exactly.
Ask "how likely is this requirement to change once we start?" If the honest answer is "very," lean Agile. If the honest answer is "basically never," Waterfall’s predictability is worth its rigidity.

related topics

reference