Rough Order of Magnitude Estimation

A comprehensive guide to creating, communicating, and refining preliminary cost and effort estimates for engineering projects.

-25% to +75%
Typical ROM Variance
Class 5
AACE Estimate Classification
0–2%
Project Definition Level
Concept
Typical Project Phase

What Is a ROM?

A Rough Order of Magnitude (ROM) estimate is a high-level, preliminary approximation of the cost, effort, or duration of a project. It is produced during the earliest stages of the project lifecycle—typically during concept screening, feasibility analysis, or opportunity assessment—when very little is known about the final scope, design, or execution plan.

Unlike detailed or definitive estimates, a ROM is intentionally imprecise. Its purpose is not to set a budget but to provide decision-makers with enough information to evaluate whether a project warrants further investment of time and resources. The expected accuracy range is wide—commonly -25% to +75%, and sometimes as broad as -50% to +100% for highly uncertain or novel programs.

This variance is not a deficiency—it is a feature. A ROM explicitly communicates that uncertainty exists and that the estimate will be progressively refined as more information becomes available through engineering studies, prototyping, and requirements definition.

Why Is ROM Critical?

In engineering, committing resources to a project without a preliminary cost or effort assessment is a recipe for wasted capital and misaligned expectations. ROM estimates serve as the first quantitative checkpoint in the project pipeline. They enable:

01
Go/No-Go Decisions
Leadership can quickly evaluate whether a project's expected cost fits within strategic budgets before investing in detailed design work.
02
Portfolio Prioritization
When competing initiatives vie for limited capital, ROMs provide the apples-to-apples comparison needed to allocate funds to the highest-value opportunities.
03
Risk Identification
The process of building a ROM forces teams to surface unknowns, technical risks, and scope gaps early—when they are cheapest to address.
04
Stakeholder Alignment
A shared ROM gives engineers, program managers, finance, and executives a common reference point—preventing "sticker shock" downstream.
05
Scope Framing
By quantifying the cost of scope elements early, teams can make informed tradeoff decisions about what to include or defer.
The Cardinal Rule
A ROM is not a commitment, not a budget, and not a contract price. It is a decision-support tool. Every ROM should be accompanied by explicit caveats about its accuracy range, the assumptions that underpin it, and the conditions under which it will be refined. Teams that treat a ROM as a firm number will almost inevitably face cost overruns, schedule delays, and eroded trust with stakeholders.

ROM Variance Quick Reference

±25–50%
HIGH CONFIDENCE
Strong historical analogues, mature technology, well-defined scope elements
±50–75%
MEDIUM CONFIDENCE
Some precedent exists, moderate unknowns, partial scope definition
±75–100%
LOW CONFIDENCE
Novel technology, undefined scope, no comparable projects, early concept

Estimate Classification System

Understanding where ROM fits in the broader landscape of cost estimation. Based on the AACE International Recommended Practice 18R-97 classification system.

Accuracy Range by Estimate Class

Class 5 — ROM / Order of Magnitude -20% to -50% / +30% to +100%
WIDEST RANGE
Class 4 — Study / Feasibility -15% to -30% / +20% to +50%
WIDE RANGE
Class 3 — Budget Authorization -10% to -20% / +10% to +30%
MODERATE
Class 2 — Control / Bid -5% to -15% / +5% to +20%
NARROW
Class 1 — Check / Definitive -3% to -10% / +3% to +15%
TIGHT
Class Name Project Definition Typical Purpose Methodology Accuracy (Low/High)
5 ROM / Concept Screening 0%–2% Go/no-go, portfolio planning, concept feasibility Analogous, expert judgment, capacity-factored -20% to -50% / +30% to +100%
4 Study / Feasibility 1%–15% Feasibility studies, concept evaluation, preliminary budget Parametric models, equipment-factored -15% to -30% / +20% to +50%
3 Budget Authorization 10%–40% Budget approval, funding authorization, design selection Semi-detailed unit costs, assembly-level -10% to -20% / +10% to +30%
2 Control / Bid 30%–75% Cost control, bidding, change-order pricing Detailed unit costs, forced takeoffs -5% to -15% / +5% to +20%
1 Check / Definitive 65%–100% Final check estimate, schedule/cost verification Detailed takeoff, vendor quotes, subcontracts -3% to -10% / +3% to +15%
Key Insight: Estimate accuracy is a function of project definition, not estimator skill. Even the best estimator cannot produce a Class 1 result when only 2% of the project is defined. This is why estimate classification matters—it sets the right expectations for accuracy and prevents misuse of early-stage estimates as budget commitments.

Estimation Methods

Each method has distinct strengths and trade-offs. The most reliable ROMs typically combine multiple methods and cross-validate results.

MethodSpeedAccuracyData RequiredBest When...
AnalogousFastMed–HighHistorical project dataSimilar past projects exist with known costs
Expert JudgmentVery FastVariableAccess to SMEsNovel projects with no direct precedent
ParametricMediumHighStatistical models, unit ratesWell-defined parameters and reliable unit cost data
Three-Point (PERT)FastMediumO/M/P estimatesRisk quantification is important
Capacity FactoredFastMed–LowKnown facility costs + scaling factorsScaling up/down from known facility or system
METHOD 1

Analogous Estimating (Top-Down)

Analogous estimating uses the actual costs of previous, similar projects as the basis for the current estimate. The historical cost is then adjusted for differences in size, complexity, technology, location, and time (inflation).

Example: Your team built an automated optical inspection (AOI) station for PCB assemblies 18 months ago at a total cost of $285K. A new project requires a similar AOI station but with additional thermal imaging capability. Using the previous station as an analogue and applying a 1.3x complexity factor plus 4% inflation, the ROM would be approximately $285K × 1.3 × 1.04 ≈ $385K.

Strengths: Fast, grounded in real data, easy to explain to stakeholders. Weaknesses: Accuracy depends heavily on true comparability; no two projects are identical. Always document the specific adjustments you've applied and why.

METHOD 2

Expert Judgment (Delphi Method)

Expert judgment relies on the knowledge and experience of SMEs to provide estimates. The Delphi method structures this by having multiple experts provide estimates independently, then sharing anonymous results and iterating until convergence.

Example: A company wants to develop a new type of environmental test chamber with no direct precedent. Three senior engineers independently estimate the effort: Engineer A says $420K, Engineer B says $380K, Engineer C says $550K. After discussion and adjustment, the team converges on a ROM of $440K ± 40%.

Strengths: Works when no historical data exists, captures tacit knowledge. Weaknesses: Subject to anchoring bias, optimism bias, and groupthink. Mitigate by using anonymous estimation rounds and including experts with diverse perspectives.

METHOD 3

Parametric Estimating

Parametric estimating uses statistical relationships between historical data and project variables. A cost-estimating relationship (CER) is developed: Cost = f(parameter₁, parameter₂, ...). Common parameters include weight, power output, channel count, or lines of code.

Example: Based on historical data, your organization has established that custom EGSE power supplies cost approximately $850 per output channel, plus a $45K base for chassis, control systems, and integration. A new project needs a 24-channel supply: $45K + (24 × $850) = $65.4K ROM.

Strengths: Objective, repeatable, scalable. Weaknesses: Requires robust historical data to build valid CERs. The relationship must hold across the range being estimated—extrapolating outside the data range is risky.

METHOD 4

Three-Point Estimating (PERT)

Instead of a single point estimate, three values are gathered: Optimistic (O), Most Likely (M), and Pessimistic (P). These are combined using one of two formulas:

TRIANGULAR
(O + M + P) ÷ 3
Equal weight to all three
PERT / BETA
(O + 4M + P) ÷ 6
Weights "most likely" 4×

The PERT formula also gives you standard deviation: σ = (P − O) ÷ 6, enabling probability-based range statements. For example, the PERT estimate ± 2σ captures approximately 95% of likely outcomes.

Example: For a cable harness manufacturing project: O = $18K, M = $25K, P = $45K. PERT = (18 + 100 + 45) ÷ 6 = $27.2K. σ = (45 − 18) ÷ 6 = $4.5K. The 95% confidence range is $18.2K – $36.2K.
METHOD 5

Capacity-Factored (Power Law Scaling)

Used extensively in process industries and large-scale systems, this method scales cost from a known reference using a power law relationship:

Cost₂ = Cost₁ × (Capacity₂ / Capacity₁)n

The exponent n (often called the "six-tenths factor" since 0.6 is commonly used) reflects economies of scale—doubling capacity doesn't double cost. Different industries and equipment types have established scaling factors based on empirical data.

Key Factors in ROM Development

A reliable ROM accounts for these critical variables. Neglecting any of them can introduce significant bias or inaccuracy.

Scope Definition Level

Critical

The single biggest driver of ROM accuracy. At the ROM stage, scope is inherently incomplete—accept this and document what is included, what is excluded, and what remains undefined. A clearly bounded incomplete scope is far more useful than an ambitiously detailed but fragile one. State assumptions explicitly: "This ROM assumes a single-string architecture with no redundancy" is infinitely more useful than silence.

Complexity & Novelty

Critical

Novel or first-of-its-kind work demands wider variance bands. If your project involves unproven technology, new manufacturing processes, or integration patterns that haven't been attempted before, your ROM range should reflect that uncertainty. A common rule of thumb: if less than 70% of the work has been done before, consider widening the range by an additional 25–50%.

Risk & Contingency

Critical

Contingency is not padding—it is a quantified allocation for known unknowns. At the ROM stage, contingency should typically be 25–100% depending on project maturity. Separate management reserve (for unknown unknowns) from contingency (for identified risks). Document the rationale: "30% contingency applied to account for immature vendor supply chain and untested integration approach."

Technology Readiness Level

Important

TRL provides a standardized framework for assessing technology maturity. Lower TRLs (1–4) indicate basic research through lab validation—these carry significantly more cost risk than flight-proven technologies (TRL 7–9). Map each major subsystem to a TRL and adjust your ROM ranges accordingly. A system with a TRL-3 critical component should have a wider overall range than one built entirely on TRL-7+ elements.

Resource Availability

Important

Availability and cost of skilled labor, specialized equipment, and test facilities directly impact project costs. If your ROM assumes in-house fabrication but the shop is at 95% capacity, you may need to factor in outsourcing premiums or schedule delays. Account for loaded labor rates (salary + benefits + overhead) rather than base salary alone.

Regulatory & Compliance

Important

Regulatory requirements can add 10–40% to project costs through certification testing, documentation, quality assurance processes, and design constraints. For aerospace, defense, or medical device projects, compliance activities (DO-178C, MIL-STD, IEC 62304) are not optional add-ons—they are fundamental scope elements that must be included in the ROM.

Market Conditions & Inflation

Consider

For projects spanning more than 12 months, material and labor cost escalation can be significant. Apply escalation factors based on producer price indices for relevant commodity categories. Supply chain disruptions, tariffs, and component obsolescence can introduce cost spikes that should be acknowledged in your assumptions.

Stakeholder Expectations

Consider

Understand what decision the ROM will support and calibrate accordingly. A ROM for an internal R&D prioritization meeting needs different presentation than one for a customer proposal. Always lead with the range and caveats, not the point estimate. Frame it as: "We expect costs between X and Y, depending on [key variables]."

Step-by-Step ROM Process

Follow this systematic 8-step process. Check off each step as you complete it—your progress is saved automatically.

0 / 8
1

Define the Problem or Opportunity

Clearly articulate what the project aims to achieve, even if the specific solution is undefined. Capture the business driver, the mission need, and the key success criteria. This framing prevents the ROM from drifting in scope as different stakeholders add requirements.

2

Identify Key Deliverables & Work Packages

Break the project into major phases or deliverables at a high level. For hardware projects, this might include: design/engineering, procurement, fabrication, integration & test, documentation, and project management. Even at ROM level, this decomposition improves accuracy by preventing omissions.

3

Gather Analogous Data & Expert Input

Search your organization's project archives for analogous work. Consult SMEs with relevant experience. Use the Delphi method for controversial or novel estimates. Cross-reference multiple data sources to identify outliers and build confidence in the central estimate.

4

Document All Assumptions

This is paramount. Every assumption—technical, commercial, schedule, resource—must be written down. Assumptions about component availability, team composition, test facility access, subcontractor rates, and design approach all materially affect the estimate. An undocumented assumption is a hidden risk.

5

Apply Estimation Methods

Select the most appropriate method(s) based on available data. For the most robust ROM, apply two or more methods independently and compare results. Significant divergence between methods signals areas that need deeper investigation or wider uncertainty bands.

6

Calculate a Range (Not a Point)

Always express your ROM as a range. Use three-point estimating where possible. The range communicates uncertainty honestly and prevents premature anchoring on a single number. Consider expressing the range asymmetrically if downside risk is greater than upside (e.g., -20% / +60%).

7

Add Contingency & Management Reserve

Apply contingency for known risks (typically 25–50% for well-understood work, 50–100% for novel work). Consider a separate management reserve for unknown unknowns. Document how the contingency was derived—whether by risk register analysis, historical overrun data, or expert judgment.

8

Communicate with Caveats

Present the ROM with clear labeling: state it is a Class 5 estimate, the accuracy range, key assumptions, major risks, and the plan for progressive refinement. Include a "basis of estimate" document that traces each cost element back to its source. Make clear that the ROM will be updated as the project matures.

Common Pitfalls

These are the most frequently observed mistakes in ROM estimation. Being aware of them dramatically improves estimate quality.

CRITICAL Treating a ROM as a Firm Budget

The single most damaging mistake. When a ROM is locked into a baseline budget, the project is set up for failure from day one. A $500K ROM with +75% variance means the actual cost could legitimately reach $875K—if stakeholders have mentally committed to $500K, the project will be perceived as "over budget" even when the final cost was always within the stated range. Always communicate the ROM as a range and ensure decision-makers understand the accuracy class.

CRITICAL Anchoring Bias & Optimism Bias

Anchoring occurs when the first number mentioned dominates subsequent estimates—if a manager says "I think this is about a $200K project," subsequent estimates will cluster around $200K regardless of the actual evidence. Optimism bias causes estimators to systematically underweight negative outcomes. Research by Bent Flyvbjerg shows that large engineering projects overrun initial estimates by an average of 28%. Mitigate by using blind estimation rounds, requiring documented rationale for each estimate, and applying reference class forecasting.

HIGH Insufficient Contingency

Teams often underallocate contingency because it "makes the number too high." But an honest ROM with adequate contingency is far more valuable than an artificially low number that will inevitably be exceeded. A 10% contingency on a Class 5 estimate is almost certainly insufficient. For novel engineering work, 50–100% contingency is appropriate and defensible.

HIGH Single Point Estimates

A single number ("The project will cost $350K") implies a precision that doesn't exist at the ROM stage. It creates false confidence and removes the ability to have a nuanced discussion about risk. Always present a range, and whenever possible, describe the probability distribution—"We are 50% confident costs will fall between $280K and $400K, and 90% confident they will fall between $220K and $520K."

HIGH Undocumented Assumptions

Without explicit assumptions, no one can evaluate whether the ROM is reasonable, and it cannot be properly adjusted when conditions change. If your ROM assumes specific labor rates, vendor lead times, design reuse percentages, or test facility availability, these must be documented. When an assumption proves wrong, the ROM can be systematically updated rather than guessed at.

MODERATE Ignoring Historical Data

Organizations that don't capture and curate project cost data are doomed to repeat estimation errors. Build a lessons-learned database that includes actual vs. estimated costs, the root causes of variances, and the conditions that made analogies valid or invalid. This institutional knowledge is one of the most valuable assets an engineering organization can develop.

MODERATE Wrong Experts at the Table

A mechanical design lead estimating software integration effort, or a software architect estimating PCB fabrication costs, will produce unreliable numbers. Ensure each major cost element is estimated by someone with direct, recent experience in that domain. Cross-functional review catches the gaps that single-discipline estimation misses.

Real-World Engineering Examples

Illustrative scenarios showing how ROM estimation applies across different engineering contexts.

SCENARIO 1

Custom EGSE Power Supply System

Context: A spacecraft program needs a custom electrical ground support equipment (EGSE) power supply rack to simulate the spacecraft's power bus during integration and test. The system requires 16 programmable power channels, safety interlocks, and a custom software control interface.

Method Used: Parametric + Analogous. The team referenced a similar 12-channel system built last year ($180K total) and applied parametric scaling for the additional channels plus a complexity factor for the new software interface.

ROM Breakdown:

Hardware (COTS supplies) ... $48K
Custom chassis/wiring .... $32K
Control software ......... $55K
Safety interlock design ... $18K
Integration & test ....... $28K
Documentation ............ $12K
Project management ....... $22K
──────────────────────────────
Subtotal ................ $215K
Contingency (35%) ....... $75K
ROM Total ............... $290K
Range: $215K – $390K (±35%)
SCENARIO 2

Automated Test Station Upgrade

Context: An existing functional test station for circuit card assemblies needs to be upgraded to support a new product variant. The upgrade involves new stimulus/measurement channels, updated test sequences, and a modified fixture.

Method Used: Analogous with adjustment factors. The original station build cost $120K. The upgrade is estimated at 40% of the original cost based on similar upgrade projects, plus a 20% factor for the new fixture tooling.

ROM Calculation:

Base upgrade (40% of $120K) $48K
Fixture tooling ........... $14K
Test software update ...... $18K
──────────────────────────────
Subtotal ................. $80K
Contingency (25%) ........ $20K
ROM Total ................ $100K
Range: $75K – $130K (±30%)
SCENARIO 3

Novel Environmental Monitoring System

Context: A research lab wants to develop a custom distributed sensor network for real-time environmental monitoring across a large test facility. No similar system has been built in-house. The project involves custom PCB design, embedded firmware, a wireless mesh network, and a data visualization dashboard.

Method Used: Expert judgment (Delphi) + Three-Point. Given the novelty, three senior engineers provided independent O/M/P estimates. The PERT formula was applied to the consensus values.

Three-Point Estimates:

Optimistic (O) .......... $320K
Most Likely (M) ......... $480K
Pessimistic (P) ......... $780K
──────────────────────────────
PERT = (320+1920+780)/6 = $503K
σ = (780-320)/6 = $76.7K
──────────────────────────────
Contingency (50%) ....... $252K
ROM Total ............... $755K
95% range: $350K – $657K
⚠ Low confidence — novel system

ROM Calculator

Build a work breakdown with labor hours, rates, and material costs per line item. Each item computes its own O/M/P cost range, and everything rolls up into PERT totals, σ, confidence ranges, and charts.

Project Information

Work Breakdown Structure

0 items

Monte Carlo Simulation

Run thousands of random trials across all WBS line items. Each trial samples every item independently from its PERT-Beta or Triangular distribution, then sums—capturing combined uncertainty realistically.

Add line items in the ROM Calculator, then click "Run Monte Carlo →" — or enter a single manual estimate below.

Manual Single-Item Input

Calculation History

View and manage your saved ROM estimates. Click any entry to load it into the calculator.

No saved estimates yet. Use the ROM Calculator to create your first estimate.

Glossary & Quick Reference

Key terms and abbreviations used in ROM estimation.

TermDefinition
ROMRough Order of Magnitude — a preliminary, high-level cost or effort estimate with wide accuracy variance.
PERTProgram Evaluation and Review Technique — a weighted three-point formula: (O + 4M + P) ÷ 6.
AACEAssociation for the Advancement of Cost Engineering — publisher of the widely used estimate classification system (18R-97).
TRLTechnology Readiness Level — a 1-to-9 scale assessing technology maturity, from basic principles (1) to flight-proven (9).
SMESubject Matter Expert — an individual with deep, specialized knowledge in a relevant domain.
CERCost Estimating Relationship — a mathematical expression relating cost to one or more project parameters.
ContingencyAn allocated budget reserve for identified risks and known unknowns. Not the same as management reserve.
Management ReserveA budget reserve held for unknown unknowns — risks that have not been identified or characterized.
EGSEElectrical Ground Support Equipment — custom electronics used to test, simulate, or support spacecraft and systems during ground operations.
WBSWork Breakdown Structure — a hierarchical decomposition of project deliverables into manageable work packages.
Basis of Estimate (BOE)A document that records the methodology, data sources, assumptions, and rationale behind an estimate.
Reference Class ForecastingA technique that bases predictions on actual outcomes of similar past projects, correcting for optimism bias.
Further Reading: AACE International Recommended Practice 18R-97 (Cost Estimate Classification System), PMI PMBOK Guide Chapter 7 (Cost Management), NASA Cost Estimating Handbook, GAO Cost Estimating and Assessment Guide (GAO-20-195G).