📐 PMP Estimating — Complete Exam Study Guide
Domain: Process | All Approaches (Predictive, Agile, Hybrid) | Eng. Ahmad Safi, PE
Estimating is the process of quantifying the likely amount of resources, time, or cost required to complete project work. It is NOT guessing — it is informed prediction based on data, experience, and judgment.
In the PMP context, estimating applies to three key areas:
- 🕐 Schedule — How long will activities take?
- 💰 Cost — How much money is needed?
- 👷 Resources — How many people, equipment, materials?
When Does Estimating Happen?
| Timing | Activity | Approach |
|---|---|---|
| Early (Initiation) | Order-of-magnitude estimate | Analogous or Parametric |
| Planning | Detailed budget & schedule | Bottom-Up or Three-Point |
| Executing | Revised estimates (re-estimating) | EVM, forecasting |
| Agile Sprints | Story point estimation | Planning Poker, Velocity |
Poor estimating is one of the top causes of project failure. The PMP exam expects you to recognize the consequences of inaccurate estimates and the PM's responsibility.
- Budget overruns
- Schedule delays
- Scope cuts to compensate
- Stakeholder dissatisfaction
- Team burnout
- Wasted resources (Parkinson's Law)
- Lost competitive bids
- Reduced ROI
- Slack absorbed (Student Syndrome)
- Stakeholder distrust
Estimates vary in accuracy depending on the project phase and available information.
| Estimate Type | Accuracy Range | When Used | Also Called |
|---|---|---|---|
| Rough Order of Magnitude (ROM) | -25% to +75% | Initiation / Feasibility | Ballpark estimate |
| Preliminary Estimate | -15% to +50% | Early Planning | Conceptual estimate |
| Budget Estimate | -10% to +25% | Mid-planning | Authorization estimate |
| Definitive Estimate | -5% to +10% | Late Planning | Control estimate |
| Final/Control Estimate | -3% to +5% | Execution (baseline) | Engineering estimate |
Progressive Elaboration of Estimates
As the project progresses and more information becomes available, estimates are refined. This is rolling wave planning in action.
Estimate Accuracy Increases Over Time →
PMBOK 7 shifted from process groups to performance domains and principles. Estimating now spans multiple domains:
- 📋 Planning — Creating cost and schedule baselines
- ⚙️ Project Work — Monitoring actuals vs. estimates
- 📦 Delivery — Estimating value and features (Agile)
- 📊 Measurement — EVM, forecasting, variance analysis
| Predictive (Waterfall) | Agile |
|---|---|
| ROM → Definitive → Baseline | Story Points → Velocity → Release Plan |
| Cost & Schedule Baselines | Product Backlog + Sprint Velocity |
| Bottom-up, Parametric, Three-Point | Planning Poker, Affinity, T-shirt |
| Work Breakdown Structure (WBS) | Product Backlog / User Stories |
Analogous Estimating uses historical data from similar past projects to estimate the current project. It is a form of expert judgment supported by historical information.
- Early phases (limited information)
- Quick estimate needed
- Similar past project exists
- ROM or budget estimate needed
- Less accurate than bottom-up
- Depends on quality of historical data
- Requires similar context
- Not suitable for unique projects
- "Less accurate but quick" → Analogous
- "Uses expert judgment + historical info" → Analogous
- "Top-down estimate" → Analogous
- Best in initiation or early planning
Parametric Estimating uses a statistical relationship between historical data and other variables (parameters) to calculate an estimate.
Estimate = Rate per Hour × Number of Hours
Estimate = Cost per Square Foot × Total Area
- Construction: $85/sq ft × 5,000 sq ft = $425,000
- Software: 20 hrs/function point × 500 FP = 10,000 hrs
- Bridge Inspection: $1,200/span × 18 spans = $21,600
- Training: 2 hrs/person × 50 staff = 100 hrs
- More accurate than analogous
- Scalable and repeatable
- Works well with clear metrics
- Requires reliable historical data
- Less useful for unique work
- Data quality determines accuracy
Bottom-Up Estimating involves estimating the cost/duration of individual work packages or activities and then aggregating (rolling up) to get the total project estimate.
Requires a detailed WBS (Work Breakdown Structure).
| Work Package | Cost |
|---|---|
| Deck Removal | $45,000 |
| Rebar Inspection | $12,000 |
| Concrete Placement | $98,000 |
| Waterproofing | $23,000 |
| Traffic Control | $18,000 |
| Total Project Estimate | $196,000 |
- "Most accurate estimate" → Bottom-Up
- "Requires WBS / work packages" → Bottom-Up
- "Time-consuming but thorough" → Bottom-Up
- Best in detailed planning phase
- "Aggregating individual estimates" → Bottom-Up
Three-Point Estimating uses three estimates to define an approximate range for an activity's duration or cost. It accounts for uncertainty and risk.
SD = (P – O) / 6
Variance = SD²
(No weighting — all equal)
Where: O = Optimistic, M = Most Likely, P = Pessimistic
O = 4 days, M = 7 days, P = 16 days
PERT: (4 + 4×7 + 16) / 6 = (4+28+16)/6 = 48/6 = 8 days
SD: (16-4)/6 = 12/6 = 2 days
Triangular: (4+7+16)/3 = 27/3 = 9 days
✅ PERT gives 8±2 days → 68% chance between 6-10 days; 95% chance between 4-12 days (±2 SD)
| Range | Confidence |
|---|---|
| E ± 1 SD (±1σ) | 68.27% |
| E ± 2 SD (±2σ) | 95.45% |
| E ± 3 SD (±3σ) | 99.73% |
| E ± 6 SD (±6σ) | 99.9997% (Six Sigma) |
- PERT weights the most likely (×4) — use when risk matters
- Triangular = simple average — equal weight to all three
- "Accounts for uncertainty/risk" → Three-Point
- The PMP exam often gives you O, M, P and asks for PERT or SD
- PERT came from the Navy Polaris missile program (1958)
Expert Judgment involves consulting individuals or groups with specialized knowledge, training, or experience to estimate cost, duration, or resources.
Expert judgment is used as a tool in nearly every estimating technique — it is rarely used alone.
- Subject Matter Experts (SMEs) from within the team
- Consultants (e.g., licensed PEs, engineers)
- Industry groups or professional associations
- Stakeholders with domain expertise
- PMO (Project Management Office)
- Vendors and contractors
Reserve Analysis is used to determine the amount of contingency and management reserve needed to handle uncertainty in project estimates.
- Handles known unknowns (identified risks)
- Part of the cost/schedule baseline
- PM can approve use
- Example: "10% reserve for weather delays"
- Handles unknown unknowns (unidentified risks)
- NOT part of the baseline
- Requires management approval to use
- Example: Unexpected regulatory changes
Project Budget = Cost Baseline + Management Reserve
BAC (Budget at Completion) = Cost Baseline
Cost estimates must account for all types of costs in a project:
| Cost Type | Definition | Example |
|---|---|---|
| Direct Costs | Directly traceable to project | Labor, materials, equipment |
| Indirect Costs | Shared overhead not directly attributed | Rent, utilities, admin staff |
| Fixed Costs | Do not change with work volume | Equipment lease, permits |
| Variable Costs | Change with output/volume | Fuel, hourly labor, materials |
| Sunk Costs | Already spent — irrelevant to decisions | Past feasibility study cost |
| Opportunity Cost | Value of best alternative foregone | Choosing Project A over B |
- Sunk costs should NOT influence future decisions — ignore them!
- Opportunity cost → always pick the higher NPV project when choosing
- "Shared overhead allocated to project" → Indirect Cost
- Fixed costs remain constant regardless of how much work is done
Cost of Quality (CoQ) is the total cost of all efforts related to quality — both conformance and non-conformance costs.
Appraisal: Testing, inspection, audits
Money spent to avoid defects
External Failure: Warranty, liability, customer complaints
Money spent because of defects
Vendor Bid Analysis (also called Bid/Quote Analysis) examines bids and quotes from vendors to determine whether the proposed prices are reasonable and within the project budget.
- Compares multiple vendor bids for the same scope
- Identifies outliers (too high or too low)
- Considers total cost of ownership, not just purchase price
These three concepts are heavily tested on the PMP exam — many candidates confuse them.
| Term | Definition | Example |
|---|---|---|
| Effort | Total labor hours needed (person-hours) | 40 person-hours (2 workers × 20 hrs each) |
| Duration | Calendar time from start to finish | 5 days (if 2 workers each work 4 hrs/day) |
| Elapsed Time | Clock time including non-working periods | 7 calendar days (includes weekend) |
(Assumes resources work full-time on the activity)
1 person working 8 hrs/day → 80/8 = 10 days duration
2 people working 8 hrs/day → 80/(2×8) = 5 days duration
But adding resources doesn't always cut duration proportionally → Brook's Law
Resource estimating determines the types and quantities of human resources, equipment, and materials required to complete project work.
Process: Estimate Activity Resources (Schedule Management)
- What type of resource? (Skill level, specialty)
- How many? (Number of personnel, machines)
- When? (Availability, calendar)
- Where? (Location-based resource availability)
Resource Histogram
A bar chart showing resource usage over time. Helps identify resource overallocation.
| Resource Leveling | Resource Smoothing | |
|---|---|---|
| Changes critical path? | YES (may extend) | NO |
| Respects constraint? | Resource limits | Time limits (float) |
| When used? | Overallocation exists | Within float available |
- Resource leveling → may delay project (ok)
- Resource smoothing → uses float, won't delay project
- Outputs: Resource requirements, Resource breakdown structure (RBS)
Velocity is the amount of work a team completes in a sprint, measured in story points. Used in Agile to forecast future sprint capacity.
Estimated Sprints = Total Backlog Points / Average Velocity
Release Date = Today + (Estimated Sprints × Sprint Length)
Average Velocity = (22+28+25)/3 = 25 points/sprint
Remaining backlog = 200 points
Sprints remaining = 200/25 = 8 sprints
With 2-week sprints → ~16 weeks to completion
- Velocity is team-specific — not transferable across teams
- New teams → use low/conservative initial velocity
- Velocity improves over time as team matures
- "Capacity-based sprint planning" uses velocity
The Basis of Estimates (BOE) is a document that explains HOW estimates were developed — the assumptions, constraints, and methodology used.
Think of it as the "show your work" document for estimates.
- Estimating methodology used (analogous, parametric, etc.)
- Assumptions made (e.g., 8-hr workday, no overtime)
- Constraints identified
- Range of possible outcomes (±%)
- Confidence level of the estimate
- Known risks that could affect the estimate
Different sources list slightly different ranges. The PMP exam tests the general concept — not exact numbers. Know the pattern:
| Estimate Class | Accuracy | Phase | Technique |
|---|---|---|---|
| ROM (Rough Order of Magnitude) | -25% to +75% | Initiation | Analogous |
| Conceptual/Feasibility | -15% to +50% | Early Planning | Analogous/Parametric |
| Preliminary/Budget | -10% to +25% | Mid-Planning | Parametric/Bottom-Up |
| Definitive/Control | -5% to +10% | Detailed Planning | Bottom-Up |
Earned Value Management (EVM) integrates scope, schedule, and cost to measure project performance against the baseline.
| Variable | Meaning | Also Known As |
|---|---|---|
| PV | Planned Value — budgeted cost for work scheduled | BCWS |
| EV | Earned Value — budgeted cost for work performed | BCWP |
| AC | Actual Cost — actual cost for work performed | ACWP |
| BAC | Budget at Completion — total project budget | — |
| Metric | Formula | Good if... |
|---|---|---|
| Schedule Variance (SV) | EV – PV | SV > 0 (ahead) |
| Cost Variance (CV) | EV – AC | CV > 0 (under budget) |
| Schedule Performance Index (SPI) | EV / PV | SPI > 1 (ahead) |
| Cost Performance Index (CPI) | EV / AC | CPI > 1 (under budget) |
At Month 3: PV = $60,000 | EV = $48,000 | AC = $55,000
SV = 48K – 60K = –$12,000 → Behind schedule
CV = 48K – 55K = –$7,000 → Over budget
SPI = 48/60 = 0.80 → 80% of schedule efficiency
CPI = 48/55 = 0.87 → Getting $0.87 of value per $1 spent
SV = EV – PV (E–P: "Earned minus Planned")
CV = EV – AC (E–A: "Earned minus Actual")
Forecasting estimates the future cost to complete remaining work based on current performance.
| Formula | When to Use |
|---|---|
| ETC = EAC – AC | How much MORE is needed to finish |
| EAC = BAC / CPI | Current CPI will continue (most common) |
| EAC = AC + ETC (bottom-up) | Re-estimate remaining work fresh |
| EAC = AC + (BAC – EV) | Past variance was atypical; future at planned rate |
| EAC = AC + [(BAC–EV)/(CPI×SPI)] | Both cost and schedule performance matter |
TCPI(EAC) = (BAC – EV) / (EAC – AC)
If TCPI > 1 → must work MORE efficiently than current
If TCPI < 1 → can afford to be LESS efficient
EAC = 120K / 0.87 = $137,931 (most likely final cost)
ETC = 137,931 – 55,000 = $82,931 (remaining cost)
VAC = BAC – EAC = 120,000 – 137,931 = –$17,931 (over budget at completion)
Story Points are a unit of measure for expressing the overall effort or complexity required to fully implement a product backlog item (user story). They are relative, not absolute (not hours).
- Complexity of the work
- Amount of work (effort)
- Uncertainty/risk
- Technical debt potential
- Fibonacci: 1, 2, 3, 5, 8, 13, 21...
- T-Shirt: XS, S, M, L, XL
- Powers of 2: 1, 2, 4, 8, 16
- Linear: 1, 2, 3, 4, 5...
Story B (dashboard redesign) = 13 points
Story B is ~4× more complex than Story A — but NOT necessarily 4× the hours.
Planning Poker is a consensus-based estimating technique used in Agile teams. Each team member selects a card with a story point value. Cards are revealed simultaneously to prevent anchoring bias.
How It Works
- Product Owner reads a user story
- Team members ask clarifying questions
- Each member selects a card (secretly)
- All cards revealed at the same time
- Outliers explain their reasoning
- Process repeats until consensus is reached
- Planning Poker = Delphi-like technique (consensus through rounds)
- "Prevents groupthink" → Planning Poker
- Cards reveal simultaneously to avoid anchoring
- No manager/senior dominance in estimates
T-Shirt Sizing is a quick, relative estimating technique using clothing sizes (XS, S, M, L, XL, XXL) to categorize work items. Used for high-level backlog grooming.
| Size | Points Equivalent | Example Task |
|---|---|---|
| XS | 1 | Fix a typo on a webpage |
| S | 2 | Add a validation rule |
| M | 5 | Build a search filter |
| L | 13 | Redesign checkout flow |
| XL | 21+ | Integrate payment gateway |
Affinity Estimating is a fast technique where the team groups user stories into buckets (like T-shirt sizes or point groups) by similarity — without detailed discussion of each item individually.
How It Works
- Write each backlog item on a card
- Team silently sorts cards into groups by relative size
- Label groups with point values or T-shirt sizes
- Discuss outliers and re-sort as needed
Triangular = (O + M + P) / 3
SD = (P – O) / 6 | Variance = SD²
SPI = EV / PV | CPI = EV / AC
EAC = AC + (BAC – EV) (past atypical)
EAC = AC + ETC (fresh re-estimate)
ETC = EAC – AC
VAC = BAC – EAC
TCPI(BAC) = (BAC – EV) / (BAC – AC)
Velocity = Points Completed / Sprint
Sprints to Complete = Remaining Points / Velocity
Project Budget = Cost Baseline + Management Reserve
BAC = Cost Baseline (NOT project budget!)
Click any keyword below for a quick definition:
- BAC ≠ Project Budget — BAC = Cost Baseline. Management Reserve is separate.
- Sunk costs are irrelevant — never factor them into future decisions.
- Story points ≠ hours — they measure relative complexity, not time.
- ROM range is wide — -25% to +75%, not symmetrical!
- Analogous is TOP-DOWN — uses historical data from similar projects.
- Bottom-up is MOST ACCURATE — but takes the most time.
- PERT weights Most Likely 4× — Triangular gives equal weight.
- Contingency reserve ≠ Management reserve — different approval levels.
- CPI < 1 = OVER BUDGET — CPI > 1 = under budget (less intuitive!).
- Resource leveling may EXTEND the schedule — smoothing does NOT.
| Exam Says... | Answer Is... |
|---|---|
| Early estimate, limited info, quick | Analogous |
| Rate × quantity, mathematical | Parametric |
| Most accurate, detailed, WBS | Bottom-Up |
| Accounts for uncertainty/risk | Three-Point/PERT |
| Agile sprint capacity | Velocity |
| Agile consensus estimating | Planning Poker |
| 100 backlog items, quick sort | Affinity Estimating |
| Known unknowns → reserve | Contingency Reserve |
| Unknown unknowns → reserve | Management Reserve |
| CPI continues → final cost? | EAC = BAC/CPI |
| Past was atypical → final cost? | EAC = AC + (BAC–EV) |
- "BIG → BAC / small → EV–AC": Budget is Big (BAC), Earned is always in the middle (EV–PV, EV–AC)
- "P4M": PERT = (O + 4M + P) / 6 — four times Most Likely
- "SV Positive = Speedy; CV Positive = Cash saved"
- "CPI < 1 = Cost Problem" — over budget despite counterintuitive direction
- "Contingency = Known | Management = Mystery"
- "Analogous = Similar past story"
Your notes are auto-saved to this browser. Type freely.