estimation

name	estimation
description	Estimate software project effort, complexity, uncertainty, capacity, cost, and delivery ranges using scoped work, decomposed tasks, dependencies, and implementation risk.

Purpose

This skill helps forecast the effort, time, cost, capacity, and uncertainty associated with software project work.

Use this skill when:

• Sizing tasks or milestones
• Forecasting delivery ranges
• Planning capacity
• Comparing implementation options
• Evaluating complexity
• Preparing for scheduling

Inputs

Expected inputs:

• Decomposed tasks
• Scope boundaries
• Acceptance criteria
• Dependencies
• Architecture assumptions
• Team capacity
• Known risks
• Target confidence level

Inputs usually come from decomposition, dependency-analysis, and architecture. If the work cannot be sized, use decomposition and dependency-analysis first.

Estimation Principles

Always optimize for:

1. Explicit assumptions
2. Range-based estimates
3. Visible uncertainty
4. Complexity awareness
5. Capacity realism

Avoid optimizing solely for:

• False precision
• Best-case timelines
• Deadline appeasement
• Ignoring review and testing
• Treating all engineers as interchangeable

Process

Step 1: Clarify Estimation Unit

Determine whether estimating:

• Task effort
• Milestone duration
• Team capacity
• Calendar time
• Cost
• Delivery confidence

Flag estimates requested without enough scope or acceptance criteria.

---

Step 2: Size Work Items

Evaluate each item by:

• Implementation complexity
• Unknowns
• Dependencies
• Testing effort
• Review effort
• Rollout effort

Use ranges when uncertainty is meaningful.

---

Step 3: Account for Risk and Capacity

Consider:

Factor	Questions
Uncertainty	What is not yet known?
Dependency Risk	What could block progress?
Team Capacity	Who is available and when?
Review Load	What reviews or approvals are needed?
Operational Work	What support, rollout, or migration work exists?

Add buffers based on actual uncertainty, not arbitrary padding.

---

Step 4: Produce Estimate

Generate:

• Effort estimate
• Calendar estimate
• Confidence level
• Assumptions
• Risks
• Work that should be spiked before committing

Anti-Patterns

Avoid:

• Estimating unscoped work
• Providing a single precise date when uncertainty is high
• Ignoring dependencies
• Excluding testing and review
• Letting deadlines define estimates
• Hiding low confidence

Reviewer Checklist

Before finalizing estimates:

• Are assumptions explicit?
• Are ranges used where appropriate?
• Are dependencies included?
• Is capacity realistic?
• Can scheduling use the estimate?

Escalation Guidance

Escalate when:

• Required confidence is higher than evidence supports
• Deadlines conflict with estimated effort
• Capacity is insufficient
• Dependencies make estimates unreliable
• Scope must change to meet timing

Example

Input:

• Tasks: database fields, checklist API, frontend checklist, analytics event
• Capacity: one backend engineer and one frontend engineer

Expected estimate:

1. Backend: 3-5 engineering days
2. Frontend: 4-6 engineering days
3. Analytics and QA: 2-3 engineering days
4. Overall calendar range: 2-3 weeks depending on review load

Recommended next skill:

• Use scheduling to turn the estimate into an execution plan.