Precedence Diagramming Method

Schedule technique · See it on the map

Runnable here

Drawing the schedule as linked task boxes, so its order is a picture, not a memory.

When to use it

Use it once you have more than a handful of tasks with real dependencies between them and need to see the whole sequence at once — it's the representation that critical path method and schedule network analysis are both built on top of.

When to avoid it

For a short, mostly-linear list of tasks, drawing a formal network adds ceremony without adding insight — a simple ordered list does the job. And it's only worth building if you're also going to keep it updated as dependencies change; a diagram that's stale by week two is worse than no diagram, because people trust it anyway.

Steps

1. List every task as a node.
2. For each pair of tasks with a real dependency, draw an arrow and record the relationship type — finish-to-start is the default, but start-to-start, finish-to-finish and start-to-finish all occur.
3. Check the network for orphan nodes (a task with no predecessor and no successor is usually a missed dependency, not a genuinely isolated task) and for cycles, which mean two tasks were each recorded as depending on the other.
4. Use the finished network as the input to critical path method or schedule network analysis.

What it produces

• A dependency network of task nodes and typed arrows that unambiguously encodes the sequence the work has to happen in.

Common pitfalls

• Every arrow gets drawn as finish-to-start by default because it's the easiest relationship to reason about, even where the real constraint is start-to-start or finish-to-finish — this artificially lengthens the schedule.
• A dependency that's organizational preference ("we'd like to do it in this order") gets recorded as a hard technical dependency, which then blocks compression options that were actually available.
• The diagram is built once at kickoff and never touched again, so it silently diverges from how the team is actually sequencing work.

Worked example

A backyard deck build records "pour footings" -> "set posts" -> "frame joists" as finish-to-start, but "install railings" as start-to-start with "stain deck boards" plus a two-day lag, since staining can begin once railing installation is under way rather than only after it completes.

Source

• PMBOK-6 §6.3.2.1

Where it comes from: this technique is named by the PMBOK Guide, 6th edition.