SOFTWARE COST INTELLIGENCE · THOUGHT LEADERSHIP

Measure the software before funding the promise.

Government cannot manage technology value consistently if it measures money, time and outcomes but not the amount of software functionality being created or changed. Functional sizing supplies that missing denominator; parametric modelling then converts measured scope and other cost drivers into an evidence-based forecast of effort, schedule, cost and uncertainty.

Edge Agents AI · 4 August 2026 · Evidence-led article for government finance, commercial and technology leaders

Explore the savings modelDownload the UCA and ScopeMaster case study (PDF) ↓

THE SAVINGS OPPORTUNITY

Government buys software without a consistent unit of scope.

The UK public sector spends more than £26bn a year on digital and data. Yet departments cannot consistently compare the functional scope delivered for that money across programmes, suppliers or technologies. Budgets, team velocity and milestones show expenditure and activity; they do not establish how much useful software was actually specified, changed or delivered.

Functional sizing changes that. By measuring the user-visible functions a system must perform, government can create a technology-neutral baseline before investment approval or procurement, challenge supplier estimates, quantify scope growth and compare whole-life cost and quality against similar work. It will not rescue every programme, but even a small improvement across the genuinely addressable share of spend could be material: the transparent scenarios below indicate a gross annual opportunity of £104m–£520m, with £234m as the central illustration.

The practical policy opportunity is to make an ISO functional-size baseline—preferably COSMIC for its broad applicability—a required evidence item at selected Treasury, assurance and commercial gates, then validate the effect through a 12-week pathfinder before claiming cashable savings.

THE PROPOSITION

Functional sizing and parametric modelling should be a mission-critical control pair.

A functional size is an objective measure of what software must do for its users, independent of programming language, supplier, team size or delivery method. Parametric modelling uses that size alongside calibrated drivers—such as complexity, reuse, team capability, technology constraints and schedule—to forecast effort, duration, lifecycle cost and uncertainty. Used together, they allow leaders to test whether an estimate is plausible, compare delivery productivity, control scope growth and build an evidence base for future investment.

Functional size is not a target to maximise, a proxy for public value or a reason to reward more functionality. It is the measured quantity input; the parametric model is the forecasting engine. The relationship is similar to measuring floor area and specification before using a calibrated construction-cost model.

FINANCE

A comparable cost denominator

Cost per functional size unit gives CFOs a portfolio signal that can be trended through discovery, approval, delivery and live operation. Variance prompts scrutiny; it does not replace judgement.

DELIVERY

Earlier, testable forecasts

Size derived from requirements can be combined with relevant productivity evidence to forecast effort, cost and schedule—and updated as scope changes.

COMMERCIAL

Technology-neutral challenge

Buyers can compare supplier proposals and change requests on a common functional basis without prescribing a language, architecture or agile team velocity.

ASSURANCE

Visible requirements risk

Ambiguity, missing functions and unbounded scope can be found before award and quantified as uncertainty rather than hidden inside optimism or contingency.

THE STANDARDS

COSMIC first; IFPUG where appropriate.

ISO COSMIC

COSMIC measures functional user requirements through data movements and is defined by ISO/IEC 19761:2011, reconfirmed in 2025. It is open, technology-independent and applicable to business, real-time, infrastructure and modern agile software. That breadth makes it the strongest candidate for a cross-government strategic standard.

ISO/IEC 19761 · COSMIC organisation

IFPUG Function Points

IFPUG Function Point Analysis is an established ISO-recognised functional-sizing family and remains relevant, particularly for business information systems and existing benchmarks. A government policy should permit valid ISO methods while preferring COSMIC where its wider applicability and open method are advantageous.

Method selection must be declared and counts must not be mixed without a documented conversion and uncertainty treatment.

PARAMETRIC MODELLING

Measure scope, then model the delivery system.

Why the methods work together

Functional sizing answers how much software functionality is required. Parametric modelling estimates what it should take to deliver and sustain it. A defensible estimate combines functional size with calibrated technical, organisational and delivery factors, then expresses cost, effort, schedule and uncertainty as traceable outputs.

Using size without a calibrated model risks a simplistic cost-per-function calculation. Using a parametric model without an objective scope measure risks a sophisticated forecast built on an unstable input. Government should adopt the two as a joined software cost-intelligence discipline, with cross-checks against analogous projects and expert judgement.

COMMERCIAL EXAMPLE

TruePlanning from PRICE Systems / Unison

TruePlanning® is Unison's parametric cost-engineering platform. PRICE® Systems Limited and TruePlanning are part of Unison; the PRICE Systems name remains in international use. The platform combines predictive models, historical data and expert calibration to produce traceable lifecycle estimates for cost, effort, schedule and uncertainty.

TruePlanning is one commercial implementation, not the policy. A government framework should remain vendor-neutral, require transparent assumptions and calibration evidence, and permit independent challenge and alternative recognised models.

Product capabilities and corporate naming are based on Unison's published information. Any public-sector adoption would require independent evaluation and an appropriate procurement route.

PRACTICAL EVIDENCE

ENABLING TECHNOLOGY · COMMERCIAL EXAMPLE

ScopeMaster turns written requirements into measurable software scope.

ScopeMaster is an AI-assisted software requirements analysis and assurance platform. It analyses requirements and user stories for ambiguity, inconsistency, duplication and missing detail; automates functional sizing using COSMIC and IFPUG Function Points; and produces traceable evidence that can support estimation, benchmarking, procurement challenge and delivery assurance.

ScopeMaster is one possible implementation tool—not the functional-sizing standard and not the proposed policy. Cross-government adoption should remain standards-led, vendor-neutral and capable of independent verification. Departments could use ScopeMaster where it demonstrates accuracy, interoperability, security and value through competition or an appropriate procurement route.

Explore ScopeMaster Return to the Edge Agents functional-sizing article →

ScopeMaster is a commercial product. Product capabilities described here are based on ScopeMaster's published information and the supplied case study; government adoption would require independent evaluation.

From uncertain requirements to defensible cost intelligence.

The supplied case study describes how German engineering and consulting organisation IABG combined ScopeMaster and TruePlanning for a major European aerospace and defence programme.

Analyse requirementsScopeMaster derives COSMIC Function Points from the functional requirements.
Expose uncertaintyAmbiguities, inconsistencies and missing elements become visible and quantifiable.
Build the parametric modelThe functional size becomes a key input to TruePlanning alongside calibrated cost drivers, producing effort, schedule, lifecycle-cost and uncertainty estimates.
Keep traceabilityInputs and assumptions remain connected to the cost output and can be updated as requirements mature.

WHAT THE CASE SUPPORTS

Audit-ready estimation

The case supports better traceability, earlier decision support, risk visibility and repeatable cost engineering. It does not publish a cash saving or controlled comparison, so none is attributed to it here.

Download the UCA and ScopeMaster case study ↓

Case study supplied by ScopeMaster. ScopeMaster is a commercial product; COSMIC is the independent sizing standard.

INTERNATIONAL GOVERNMENT PRACTICE

This is established public-sector cost discipline, not an experimental metric.

United States

The US Government Accountability Office's Cost Estimating and Assessment Guide identifies COSMIC and function points among recognised sizing methods. It says software estimates have two basic elements—size and development effort—and stresses consistent, reproducible rules, independent control and cross-checking by more than one method.

GAO Cost Estimating Guide

Brazil

Brazil's federal administration has long used function-point measurement in software contracting. The 2010 federal ordinance describes it as an objective measure and recommends the SISP counting guide for development and maintenance procurements. Brazil demonstrates both the scalability of the approach and the need for strong governance to prevent unit pricing becoming a substitute for outcomes.

Federal ordinance

European defence context

The IABG example shows COSMIC-derived sizing feeding lifecycle cost models in a regulated multinational environment where estimates must remain explainable years later. It is directly relevant to defence, aerospace and other complex UK public programmes.

Read the UCA and ScopeMaster case study

UK opportunity

HMT and GDS found an absence of agreed digital outcome metrics, while the State of Digital Government review found no consistent digital-performance measures. Functional size can complement—not compete with—outcomes, service performance, security and user satisfaction.

Digital Spend Review

PUBLIC-PURSE OPPORTUNITY MODEL

A plausible annual opportunity: £104m–£520m.

The 2025 State of Digital Government review reports more than £26bn of annual UK public-sector digital and data spend in 2023. Not all of that is addressable by functional sizing: hardware, connectivity, commodity licences and some run costs should be excluded. Nor will every pound of better estimation become a cash saving.

£234millustrative gross annual opportunity
£7.8bnspend treated as addressable
0.90%saving as total digital spend

Formula: £26bn × 30% addressable spend × 3% improvement = £234m gross annual opportunity.

ScenarioAddressable shareImprovementGross opportunity
Conservative20%2%£104m
Central30%3%£234m
Upper illustrative40%5%£520m

These are Edge Agents scenarios, not government forecasts or ScopeMaster claims. They represent potential gross cost avoidance before implementation, training, assurance and tool costs. A cross-government pilot must establish the addressable baseline and measured counterfactual before any cashable-savings claim.

Source: State of Digital Government review, January 2025

CFO KPI DESIGN

Use a balanced control set—not one number.

MeasureDecision it supportsGuardrail
Forecast and delivered COSMIC sizeScope maturity and growthNever reward more functionality
Whole-life cost per CFPEstimate challenge and benchmarkingSegment by service type and complexity
Effort per CFP / CFP per team-monthProductivity trendCompare like with like; do not rank individuals
Defects and rework per CFPQuality and avoidable costKeep severity and user impact visible
Requirements quality and uncertaintyReadiness to fund or procureReport assumptions and confidence
Outcome and service measuresWhether the investment creates public valueFunctional size is a denominator, not the outcome

INITIAL WALKING SKELETON

Prove the control across three live projects in 12 weeks.

1. Baseline

Select three differently shaped services, capture approved requirements, current estimates, actual cost/effort and delivery outcomes.

Gate: finance, delivery and commercial owners agree the measurement boundary.

2. Size and model

Produce independent COSMIC counts, then create calibrated parametric estimates of effort, schedule, lifecycle cost and uncertainty. Compare both with existing estimates and analogous projects.

Gate: sizing repeatability, model calibration and material decision insight are demonstrated.

3. Calibrate and benchmark

Calibrate model drivers with relevant historical evidence, then compare cost, effort, duration, defects and change per CFP using comparable cohorts—not a universal league table.

Gate: the measures explain variance without driving dysfunctional behaviour.

4. Decide

Measure the pathfinder's operating cost, avoided rework, estimate variance and decisions changed; publish the method, limitations and counterfactual.

Gate: an independently reviewed business case shows benefits exceed operating cost.

CROSS-GOVERNMENT ROLLOUT ROADMAP

Move from evidence to a governed portfolio control.

Adoption should be staged, standards-led and proportionate. Functional size and calibrated parametric models become useful when the same governed evidence chain follows scope from investment case through procurement, delivery, change control and benefits review.

PHASE 1 · 0–3 MONTHS

Prove and design

Pathfinder

Run the three-project walking skeleton across different service types and delivery models.

Measurement and modelling charter

Define boundaries, COSMIC counting rules, parametric-model calibration and validation rules, uncertainty treatment, independent tolerances, data ownership and anti-gaming safeguards.

Integrated tool evaluation

Compare manual and automated sizing tools such as ScopeMaster, and parametric platforms such as TruePlanning from PRICE Systems / Unison, for accuracy, calibration, repeatability, traceability, integration, security and whole-life cost.

Decision gate: proceed only if decision value and measured benefits exceed implementation cost.

PHASE 2 · 3–9 MONTHS

Standardise at investment and commercial gates

Central guidance

GDS, GCA, HMT and IPA jointly publish proportionate guidance: where a functional-size baseline is required, permitted ISO methods and materiality thresholds.

Business cases

Add forecast functional size, calibrated parametric cost, effort and schedule ranges, uncertainty analysis and comparable evidence to selected digital spend and assurance submissions.

Procurement clauses

Require suppliers to disclose sizing assumptions, maintain traceability to requirements and report scope change in functional units without turning CFP into a crude unit-price target.

Decision gate: approve a government measurement profile and model clauses after legal, commercial and delivery review.

PHASE 3 · 9–18 MONTHS

Build capability and portfolio evidence

Independent assurance

Establish accredited counters and a separation between teams proposing scope, suppliers estimating delivery and the function assuring the count.

Delivery integration

Connect requirements repositories, Jira or Azure DevOps, commercial records and finance data so scope and cost evidence updates through delivery.

Model and benchmark service

Create a protected cross-government repository of functional sizes, actual costs, schedules and model drivers, segmented by service type, complexity and delivery context. Use completed projects to recalibrate models and publish anonymised ranges, not supplier league tables.

Decision gate: demonstrate repeatable counts, useful cohorts and no material gaming or perverse incentives.

PHASE 4 · 18–36 MONTHS

Scale proportionately across government

Portfolio KPI

CFOs and CDIOs review cost, effort, defects, rework, elapsed time and scope growth per functional unit alongside outcomes, security and user satisfaction.

Spend controls

Make the baseline mandatory only above evidenced thresholds or risk criteria, with exemptions for unsuitable commodity or predominantly non-software spend.

Continuous evaluation

Audit realised savings against a counterfactual, refresh benchmark cohorts and retire controls that cost more than the decisions they improve.

Outcome: functional size becomes a normal, auditable denominator for material custom-software investment—not a universal target or substitute for public value.

Discuss a functional-sizing pathfinder Visit ScopeMaster →