From output to outcome: the case for measurable goals
In large organisations, software development spans multiple units, departments, sub-organisations, and external contributors – each using disconnected performance measures. This fragmentation obscures end-to-end views of quality, efficiency, and delivery, pushing decisions toward intuition rather than evidence and producing inconsistent outcomes (Sürücü et al., 2020). Key Performance Indicators (KPIs) address this directly, they represent structured values enabling organisations to analyse and compare performance against specified objectives, serving as decision-grade indicators of whether engineering work improves over time and translates into business value (Bandaru, 2025).
A persistent challenge is the confusion between output and outcome. Teams may close hundreds of tickets, hit velocity targets, and maintain green dashboards while still failing to deliver meaningful results. Velocity and story points are coordination tools suited to sprint planning, not strategic indicators of business health. Elevating activity metrics to KPI status risks what Goodhart’s Law describes: when a measure becomes a target, it ceases to be a good measure (Bandaru, 2025). Metrics describe motion, while KPIs describe direction. There are four performance measure types identified:
Result Indicators
Key Result Indicators
Performance Indicators
Key Performance Indicators
With KPIs at the highest level representing core business performance tracked at regular intervals to support strategic goal health. Effective KPIs must be SMART, actionable, and balanced across financial and non-financial perspectives. Crucially, teams must understand not just what is being measured but why and how it aligns to organisational goals and vision (Smith, 2020).
Frameworks such as DORA (Digital Operational Resilience Act) demonstrate how KPI-driven approaches improve both delivery speed and system stability, two key dimensions of successful software outcomes (Daraojimba et al., 2024). Yet no single framework explains performance fully (Directio, 2025). Effective programmes complement DORA with quality, team health, and business-impact indicators, including the SPACE (Satisfaction, Performance, Activity, Communication & collaboration, and Efficiency & flow) framework for multi-dimensional developer productivity and DevEx (Developer Experience) programmes for workflow friction and enablement signals selecting a coherent set of three to seven KPIs per role, each answering a distinct strategic question (Directio, 2025; Musketeers Tech, 2026).
From Output to Outcome:
Business outcome (Goal)
Customer outcomes (What changes for users)
Product KPIs (Measure user impact)
Engineering KPIs (Enable delivery performance, e.g. DORA)
Supporting signals (NOT KPIs - diagnostic only)
Underlying work (Outputs / Activity)
KPIs as a shared language - prioritisation, efficiency, alignment
One of the most significant benefits of KPI-driven development is the common language it creates between technical teams and business stakeholders. At Volkswagen Financial Services AG (VWFS), unified KPIs were introduced explicitly to facilitate communication across organisational units, reduce coordination effort, and establish shared understanding of what measured data meant, avoiding confusion from synonymous terms (Sürücü et al., 2020). KPIs support management by providing feedback on planning and policy decisions, determining critical success factors through indicators aligned with organisational goals, when goals are expressed numerically, achievement becomes measurable and success factors identifiable (Smith, 2020).
When defined early and tied to business objectives, KPIs also underpin more effective prioritisation. Objectives should be established prior to defining KPIs. This requires clearly articulating the project’s intended outcomes, identifying metrics that meaningfully reflect progress, and ensuring that each KPI is explicitly aligned with broader business objectives. Such an approach helps prevent the adoption of measures that are easy to track but lack substantive relevance to organisational performance (BIT Studios, 2024). At VWFS, the Delay of Requirements (DoR) KPI enabled project leads to identify bottlenecks and assign resources proactively to overdue items, while the Number of Change Requests after Deadline (NoCRaD) KPI detected post-deadline modifications and enabled earlier risk interventions (Sürücü et al., 2020). More broadly, monitoring indicators across the development lifecycle helps teams identify bottlenecks, reduce failure rates, and improve deployment frequency, directly serving business and customer goals (Daraojimba et al., 2024).
The alignment function extends to user experience. UX (User Experience) KPIs derived from software requirements using machine learning can be evaluated before prototyping, enabling engineers to address UX weaknesses at the earliest lifecycle stage, reduce rework, and ensure UX compliance before implementation, preventing the omission of user experience and inadequate resource planning that absent UX requirements commonly cause (Atoum, 2023; Jain et al., 2025).
Accountability, implementation, and lessons from practice
KPI-driven development strengthens accountability across teams and management levels by creating a shared basis for evaluating decisions and tracking interventions. At VWFS, the Priority Defects per Person Day (PDpPD) KPI related higher-priority defects to development effort, enabling cross-release and cross-system comparisons regardless of size and supporting data-driven transformation decisions. Used alongside Automated Test Progress (ATP), Classification of Defects (CoD), and Manual Test Progress and Interruptions (MTPaI), it increased defect-cause transparency and drove concrete improvements, including automated regression testing and new coordination practices across development, incident management, testing, and business departments (Sürücü et al., 2020). Effective dashboards support this by delivering real-time data collection, shared visibility, and immediate bottleneck detection, though they must be built around the right questions, not merely designed to generate confidence (Bandaru, 2025; Directio, 2025).
For KPIs to fulfil their accountability function, they must be accepted and understood by all relevant stakeholders, introducing them is not enough. This requires management commitment, clear communication of what each KPI measures and feedback loops showing that input is acted upon, transforming KPIs from top-down reporting requirements into shared steering instruments. Implementation evidence points to several practical principles. Keep the KPI set small, VWFS learned this the hard way when their initial ~25 indicators proved unmanageable and most guidance settles on three to seven per role (Sürücü et al., 2020; Directio, 2025). Define each KPI before data collection starts, with a clear answer to “What decision will this number help us make?” (Musketeers Tech, 2026). Start small and iterate, VWFS began with just four indicators before expanding, and that gradual approach was key to their programme taking hold (Sürücü et al., 2020). Overall, maintain a business case that weighs the costs and benefits of each KPI, so evaluations rest on evidence rather than assumption. KPIs must always be reviewed as trends against a baseline, not point-in-time snapshots, only sustained trends allow teams to distinguish real improvement from noise (Musketeers Tech, 2026; (LinearB, 2025).
Conclusion
KPI-driven development is a strategic discipline, not a reporting exercise, connecting daily engineering work to outcomes organisations care about. When well-chosen, broadly understood, and reviewed as trends, KPIs bridge business and technology, surface hidden inefficiencies, and enable continuous improvement. Evidence from industrial implementations, literature reviews, and practitioner frameworks consistently shows that measurable goals improve focus, accountability, results and that output-focused development frequently misses real impact. Organisations embedding KPI thinking from requirements through. Release is better positioned to deliver software that genuinely serves their strategic goals.
Are your teams measuring activity or actual outcomes?
Let’s identify the KPIs that drive your software performance!
