Value & Impact
Delivery efficiency measurement (DVMS), impact KPIs, benefits realisation, and industry benchmarking.
Overview
Government ICT governance watches what goes in. Labour rates, contract pricing, compliance process. It rarely asks the question that actually matters, which is what came out the other end and whether it made any difference.
The numbers say that is the wrong way round. ADDA benchmarking across government programmes finds labour rates barely move across the market. Cost per feature delivered to production varies up to fifty-fold. Negotiating an hourly rate down 5% is not where the money is. How efficiently those hours turn into working capability is.
We work both sides of that. Teams should know why they are building something and what success would look like, and somebody should be measuring, continuously, how much capability each dollar is actually buying.
The Value Gap
ADDA benchmarking data illustrates the scale of the opportunity:
| Procurement Model | Tech Stack | Cost per Feature |
|---|---|---|
| Big vendor, fixed price | Proprietary | ~$1,000,000 |
| Timesheets, contractor skills | Proprietary | ~$100,000–$150,000 |
| Feature-based, outcomes-focused | Open source | ~$20,000–$25,000 |
None of that spread is explained by what people charge per hour. It comes from architectural complexity, how dependencies are managed, the procurement model, the technology choices, and how the team works.
How We Deliver
Impact KPIs
KPIs get set at the start of an engagement, tied to the client’s goals and to what users are supposed to get out of it. They are not an evaluation exercise bolted on at the end. The delivery team can see them throughout, and uses them to decide things.
A usable KPI is quantifiable, measures an effect on users rather than team activity, has a baseline from current performance, and has a threshold. We review them at each milestone, and change them when the early evidence shows we are measuring the wrong thing.
DVMS — Delivery Value Measurement Standard
DVMS provides continuous visibility into how efficiently investment is converted into delivered capability.
Feature classification. Each feature is classified using three fields:
| Field | Purpose | Options |
|---|---|---|
| Function type | What the feature does | UI: Add, Edit, View, Delete, List, Report; Integration: API Add, Edit, View, Delete, List |
| Change type | Nature of the change | New, Update, Remove |
| Complexity | Structural difficulty | Simple, Medium, Complex |
These produce a standardised feature point value, enabling like-for-like comparison across teams, programmes, and against industry benchmarks.
Three layers of measurement:
- Delivery efficiency — cost per feature point, delivery speed, cross-programme comparison against ISBSG benchmarks. Are we building things well?
- Estimation quality — three-way comparison between calculated estimates, product owner overrides, and actuals. Are our estimates reliable?
- Benefits realisation — tiered assessment of whether delivered capability achieves its intended outcomes. Are we building the right things?
Dashboard views are tailored to each audience:
| Audience | Focus |
|---|---|
| Senior leadership | Portfolio cost per feature point, programmes above/below benchmark, cost trajectory |
| Programme manager | Cost trends by team, delivery progress vs plan, estimation accuracy |
| Team leader | Hours per feature point, sprint output, estimation reliability |
Three-Tier Benefits Realisation
Not all benefits can be reduced to dollar values. Our framework uses three tiers: directly monetised benefits (cost savings, processing time reductions), proxy-monetised benefits (risk reduction, compliance improvement), and scored strategic benefits (sovereign capability, strategic flexibility). This avoids either ignoring non-monetary benefits or assigning them fictitious dollar values.
Principles
- Measure outputs, not just inputs. A programme that only tracks spend cannot tell efficient delivery from expensive waste. Both look like money going out.
- Transparency builds improvement. We share the efficiency metrics openly. When cost per feature point starts climbing, that is a signal to go and look, not to find someone to blame.
- Aligned incentives. Where we can, we bill against outcomes or milestones rather than time, so that we are paid for what we delivered rather than how long we took.
Policy Alignment
Impact KPIs line up with the six dimensions the OECD Digital Government Policy Framework uses for evaluation.
DVMS feature points derive from ISO/IEC 20926, the international standard for functional software size, which is what makes them comparable outside our own projects. Efficiency is then benchmarked against ISBSG, the largest industry dataset on software delivery cost and productivity.
GoSource is a founding member of the Australian Digital Delivery Alliance, whose efficiency and value framework this work feeds.
Evidence
DVMS, the Delivery Value Measurement Standard, is ours. Steven Capell wrote it, and also designed the value metrics frameworks used at ABF, Home Affairs and DAFF.
On ABF trade modernisation the value metrics, clearance time, revenue compliance and strike rate, traced back to the ABF mission and became the foundation of the business case. It secured capital funding on demonstrated value well above cost.
National Parks e-ticketing repaid its build cost four times over in three months and now processes more than $10 million a year. Export Assurance analysed pain across the $80 billion agricultural export market and found only one audit in five producing a corrective action.
ADDA’s Recoding Australia names “Process Over Outcomes” as a root cause of ICT failure, and argues for exactly this kind of measurement.
Tools & Technologies
- Feature Point Classification: Azure DevOps custom fields, Jira custom fields
- Benchmarking: ISBSG repository, IFPUG Function Point Analysis (ISO/IEC 20926)
- Dashboards: Power BI, Azure DevOps Analytics
- Benefits Assessment: Three-tier model (monetised, proxy-monetised, scored strategic)