Dependency Mapping

Dependency Mapping

Create clear visibility into critical path, cross-team blockers, and sequencing constraints.

Context

You are helping a tech lead or multiple teams understand dependencies in a complex initiative. If you have a list of teams, features, or work items, use them to build a dependency graph.

Domain Context

• Theory of Constraints (Goldratt, "The Goal"): identifying and managing the critical path is how to unblock complex work
• Larson: mismanaged cross-team dependencies are the #1 source of timeline slippage
• Systems thinking: dependencies often cascade; removing one blocker reveals the next

Key principles:

1. Explicit beats implicit: Dependencies that are unspoken get discovered painfully. Write them down
2. Differentiate types: Blocking (task B cannot start until A finishes), informational (good to know), and resource (same person needed on both tasks)
3. Shorten critical path: Focus optimization on the longest dependency chain, not everything

Instructions

1. List all work items: From your WBS or roadmap, create a flat list of epics, features, or projects with owners
2. For each item, ask: "What must be done before we can start this?" Create pairs (A depends on B)
3. Classify dependencies:
   • Blocking: Task cannot start until dependency finishes (e.g., backend API must exist before mobile integrates)
   • Resource: Same engineer needed on both (cross-team pair programming or handoff)
   • Informational: Useful to know but not blocking (e.g., design feedback for next phase)
4. Build the graph: Use a simple table or ASCII diagram. Example:

   Item                    | Depends On            | Type      | Owner
   ─────────────────────────────────────────────────────────────────
   Mobile OIDC integration | OIDC API staging      | Blocking  | Mobile Lead
   OIDC API staging        | Identity svc deployed | Blocking  | Backend Lead
   Identity svc deployed   | Infrastructure ready  | Blocking  | Infra Lead
   Support training        | Runbook completed     | Blocking  | Platform Lead
   

5. Find the critical path: Trace the longest chain of blocking dependencies. Optimizing this path moves the overall delivery date
6. Identify parallelization: Look for non-blocking items that can run in parallel (e.g., web client development and mobile can run simultaneously after API is ready)
7. Plan handoff points: For each dependency, schedule a handoff meeting. Ensure receiver knows acceptance criteria and ETA
8. Build contingency plans: For each critical path item, ask "what if this slips by 1 week?" Plan alternate routes or parallel work

Anti-Patterns

• Assuming hidden dependencies will emerge naturally: LLMs often assume teams will "figure it out" during execution. Explicit dependency mapping prevents surprises. Always document assumptions upfront.
• Ignoring resource conflicts: Two critical path items might need the same specialist. LLMs often miss these shared resource constraints. Ask "who is doing this?" for each item.
• Treating all dependencies equally: Blocking dependencies matter; informational ones are noise. LLMs sometimes create dependency graphs that are too detailed and lose signal. Focus on blocking dependencies only.
• No contingency for critical path slippage: Plans assume everything on the critical path finishes on time. Have a plan B if any critical item slips.

Further Reading

• Goldratt, Eliyahu. "The Goal: A Process of Ongoing Improvement." North River Press, 1984. Theory of Constraints.
• Larson, Will. "An Elegant Puzzle." Chapter on managing complex projects and dependencies.
• Reilly, Tanya. "The Staff Engineer's Path." Chapter on managing large organizational changes.