Analogous vs Parametric vs Bottom-Up Estimating

Three Estimating Techniques the PMP Exam Tests Repeatedly

Estimation is fundamental to project management, and the PMP exam tests three primary techniques: analogous, parametric, and bottom-up. Each operates at a different level of detail, requires different inputs, and produces estimates of different accuracy. PMP questions present scenarios and ask which technique is most appropriate — or which is being described. Knowing the trade-offs is essential.

Analogous Estimating: Expert-Based Top-Down

Analogous estimating uses historical data from similar past projects to estimate the current project. It is a top-down approach — the estimate starts at the project level and may or may not be decomposed to lower levels.

How It Works

A project manager looks at a completed project with similar characteristics and uses its actual cost or duration as the basis for the current estimate, adjusted for known differences. For example, "The last office renovation cost $2 million and took 8 months. This office is 20% larger, so we estimate $2.4 million and 9.5 months."

Characteristics

• Speed: Fastest of the three techniques.
• Cost: Least expensive to perform.
• Accuracy: Least accurate — relies heavily on the similarity between the past and current projects.
• Data required: Historical project data and expert judgment.
• Best for: Early-phase estimates when detailed scope is not yet available, or when time and budget for estimation are limited.

Parametric Estimating: Algorithm-Based Modeling

Parametric estimating uses statistical relationships between historical data and project variables to calculate estimates. It applies a mathematical model — a parameter — to a measurable quantity.

How It Works

The estimator identifies a measurable unit of work and applies a known rate. For example, "Laying cable costs $50 per meter. The project requires 10,000 meters. Estimated cost: $500,000." Or, "Each software module takes 40 hours to develop. The project has 25 modules. Estimated effort: 1,000 hours."

Characteristics

• Speed: Moderate — faster than bottom-up, slower than analogous.
• Cost: Moderate.
• Accuracy: Can be very accurate when the model is valid and the data is reliable. Accuracy depends on the quality of the underlying statistical relationship.
• Data required: A reliable model, historical rates, and accurate quantity measurements.
• Best for: Projects with repeatable, measurable units of work and reliable historical rate data.

Bottom-Up Estimating: Detailed Decomposition

Bottom-up estimating builds the project estimate from the lowest level of the WBS. Each work package or activity is estimated individually, and the estimates are aggregated upward to produce control account, phase, and project-level estimates.

How It Works

The team estimates the cost and duration of every work package. These estimates are summed to produce the total project estimate. For example, a software project might estimate each of 200 user stories individually, then sum them for the total.

Characteristics

• Speed: Slowest — requires detailed decomposition and estimation of every component.
• Cost: Most expensive to perform.
• Accuracy: Most accurate — granular estimates reduce the risk of missing work or underestimating individual components.
• Data required: A complete WBS with defined work packages, plus detailed knowledge of each component.
• Best for: Later planning phases when detailed scope is available, or when a high-confidence estimate is required for budgeting or contract negotiations.

Comparison Table for PMP Questions

• Least time/cost to produce: Analogous
• Most time/cost to produce: Bottom-up
• Least accurate: Analogous
• Most accurate: Bottom-up
• Uses historical data from similar projects: Analogous
• Uses mathematical models and rates: Parametric
• Uses detailed WBS decomposition: Bottom-up

PMP Exam Decision Guide

1. Early in the project with limited scope detail? → Analogous
2. Repeatable work units with reliable rate data? → Parametric
3. Detailed WBS available and high accuracy needed? → Bottom-up

Three-Point Estimating: A Complementary Technique

Three-point estimating (PERT) can be combined with any of the above techniques. It uses optimistic, most likely, and pessimistic estimates to account for uncertainty. The triangular estimate averages all three; the beta (PERT) distribution weights the most likely estimate four times: (O + 4M + P) / 6. Three-point estimating adds a risk dimension to whichever base technique you use.

Practice identifying estimating techniques in scenario questions. Visit our cost management topics or explore our plans for access to hundreds of PMP practice questions.