HKSM

A quantitative estimating technique that uses unit rates or statistical relationships to calculate cost or duration based on measured quantities. It multiplies a validated rate or model by the quantity of work and adjusts for known factors and risks.

Key Points

• Uses measurable quantities and unit rates or formulas to estimate cost or duration quickly and consistently.
• Accuracy depends on data quality, similarity to past work, and how well the model reflects current conditions.
• Works at high level or detailed levels and can be linear or non-linear with scaling factors.
• Rates and models should be calibrated using historical data and adjusted for complexity, location, and productivity.
• Include ranges and confidence; can be combined with risk analysis techniques such as three-point estimating or Monte Carlo.
• Useful for comparing alternatives and for updating forecasts as quantities change.

When to Use

• When scope can be expressed in repeatable, countable units (e.g., meters, features, transactions, components).
• When you have reliable historical rates, benchmarks, or a calibrated model.
• When you need fast, scalable estimates across many similar items.
• In both predictive and adaptive life cycles to forecast cost or time based on throughput or velocity.
• When seeking consistent estimates across vendors, teams, or locations.

How to Estimate

• Define measurable units for each scope element and quantify them.
• Select or build an appropriate model (unit rate, regression, productivity curve) from historical data.
• Adjust rates for context factors such as complexity, team capability, location, and calendar efficiency.
• Compute estimates by multiplying rate by quantity and adding fixed components and overhead.
• Aggregate results, add contingency based on risk, and express an estimate range and confidence.
• Validate against other techniques (analogous or bottom-up) and document the basis of estimate.

Inputs Needed

• Decomposed scope with measurable quantities or throughput metrics.
• Historical project data, industry benchmarks, and calibrated unit rates.
• Complexity and location factors, productivity rates, and cost indices.
• Resource calendars, availability, and organizational constraints.
• Risk register items that affect rates or quantities.
• Estimating policies, templates, and lessons learned.

Outputs Produced

• Cost and/or duration estimates at activity, work package, or feature level.
• Basis of estimate documenting data sources, model parameters, and adjustments.
• Estimate ranges and confidence levels suitable for decision making.
• Assumptions, exclusions, and constraints captured for traceability.
• Updates to cost, schedule, and risk artifacts; approved estimates may inform baselines.

Assumptions

• Relationships between quantities and effort, cost, or time are stable within the modeled range.
• Historical data or benchmarks are valid for the current environment and technology.
• Scope items are sufficiently similar to past items used to derive rates.
• Resource productivity and calendars will remain within expected tolerances.
• Identified risks are covered by contingency or managed through responses.

Example

You must deliver 250 standardized units. Historical data shows 3.5 hours per unit at complexity factor 1.0. Your context indicates a 1.2 complexity factor and a fixed setup of 40 hours; effective calendar efficiency is 0.90.

• Adjusted rate = 3.5 h/unit x 1.2 = 4.2 h/unit.
• Total effort before calendar = (4.2 x 250) + 40 = 1,090 hours.
• Calendar-adjusted effort = 1,090 / 0.90 = 1,211 hours (rounded).
• If labor cost is 70 per hour, cost estimate = 1,211 x 70 = 84,770.
• Add 10% contingency for known risks: total = 93,247.

Pitfalls

• Using outdated or non-comparable data without calibration.
• Ignoring fixed costs or learning curves and assuming strict linearity.
• Failing to adjust for context factors such as team capability or location.
• Double-counting contingency or applying it inconsistently across items.
• Not validating model results with another technique or expert review.
• Overconfidence in a single-point estimate without a stated range and confidence.

PMP Example Question

A project manager has reliable historical data showing the average cost per unit for installing standardized components. The scope defines exact quantities for this work. Which estimating approach should the manager use to produce a quick, consistent estimate?

1. Expert judgment from subject matter experts.
2. Bottom-up estimation for each component individually.
3. Parametric estimating using cost per unit.
4. Analogous estimating based on a past project total.

Correct Answer: C — Parametric estimating using cost per unit.

Explanation: With known quantities and validated unit rates, parametric estimating provides a fast and repeatable estimate; other methods are slower or less precise for this context.