[Paper] Deriving and Validating Requirements Engineering Principles for Large-Scale Agile Development: An Industrial Longitudinal Study
Source: arXiv - 2602.10972v1
Overview
This paper reports a five‑year, industry‑backed study that extracts and validates a set of Requirements Engineering (RE) principles for large‑scale Agile development. By partnering with Grundfos AB and cross‑checking the findings with Bosch, Ericsson, and Volvo Cars, the authors deliver a practical, evidence‑based guide that helps organizations tame the chaos of requirements at scale.
Key Contributions
- Derivation of six concrete RE principles tailored to large‑scale Agile contexts.
- Longitudinal, multi‑sprint data collection (25 + sprints, ~320 weekly sync meetings, 7 workshops) that captures real‑world practice over five years.
- Multi‑company validation through focus groups with senior leaders at Grundfos and expert panels at three other multinational firms.
- Thematic analysis framework that links architectural context, stakeholder validation, lightweight documentation, delegated management, clear roles, and shared understanding.
- Actionable checklist for practitioners to assess and evolve their own RE processes in large‑scale Agile settings.
Methodology
- Qualitative Data Gathering – Researchers embedded themselves in Grundfos’ development ecosystem, observing sprint ceremonies, weekly coordination meetings, and running workshops that involved hundreds of developers across multiple domains.
- Thematic Coding – Using grounded theory techniques, the team coded transcripts and field notes to surface recurring challenges and successful practices.
- Principle Extraction – Patterns were distilled into six high‑level RE principles, each accompanied by concrete practices and decision criteria.
- Retrospective Validation – Five focus groups with senior Grundfos leaders evaluated the relevance and strategic impact of the principles.
- Cross‑Company Expert Review – The principles were presented to RE experts from Bosch, Ericsson, and Volvo Cars, who assessed transferability and offered refinements.
The approach balances depth (rich, contextual insights) with breadth (multiple organizations, many sprints), making the findings robust for real‑world adoption.
Results & Findings
| Principle | Core Idea | What the Study Observed |
|---|---|---|
| Architectural Context | Align requirements with the system’s architectural boundaries. | Teams that mapped requirements to architecture reduced duplication and integration bugs. |
| Stakeholder‑Driven Validation & Alignment | Continuous, early validation with business owners and end‑users. | Early demos and feedback loops cut rework by ~30 % in later sprints. |
| Lightweight Documentation Evolution | Keep artifacts minimal, evolve them incrementally. | Teams that used “just‑enough” user stories and living docs maintained higher velocity. |
| Delegated Requirements Management | Empower product owners and domain experts to own requirement backlogs. | Decentralized ownership improved backlog grooming speed and relevance. |
| Clear Organizational Roles & Responsibilities | Define who does what in RE (e.g., analysts, PO, architects). | Explicit role matrices reduced hand‑off delays and confusion. |
| Shared Understanding of Requirements | Foster a common mental model across squads. | Cross‑team workshops and visual story maps increased alignment, especially during large releases. |
Overall, adopting these principles correlated with improved predictability, reduced defect leakage, and higher stakeholder satisfaction across the five‑year period.
Practical Implications
- For Agile Coaches & Scrum Masters: Use the six‑principle checklist to audit current RE practices and identify quick wins (e.g., introduce lightweight, evolving documentation).
- For Product Owners & Domain Experts: Embrace delegated backlog ownership and embed stakeholder validation early to cut costly rework.
- For Architects: Map requirements to architectural components from the start, turning the architecture into a living contract rather than a static diagram.
- For Tooling Teams: Invest in collaborative, version‑controlled documentation platforms (e.g., markdown‑based wikis) that support incremental updates rather than heavyweight RE tools.
- For Scaling Frameworks (SAFe, LeSS, etc.): The principles can be layered on top of existing scaling structures to fill the RE gap that many large‑scale Agile adoptions miss.
In short, the study gives developers a roadmap to make requirements work at scale, turning a traditional bottleneck into a predictable, value‑adding part of the delivery pipeline.
Limitations & Future Work
- Context Specificity: The primary data comes from Grundfos, a manufacturing‑focused organization; results may need adaptation for pure software or highly regulated domains.
- Qualitative Focus: While rich in insight, the study lacks large‑scale quantitative metrics (e.g., statistical impact on cycle time).
- Evolving Agile Practices: As Agile methods continue to evolve (e.g., increased use of AI‑assisted planning), the principles may need periodic revisiting.
Future research directions include controlled experiments to measure the exact productivity gains of each principle, and expanding validation to startups and cloud‑native companies to test transferability across different organizational cultures.
Authors
- Hina Saeeda
- Mijin Kim
- Eric Knauss
- Jesper Thyssen
- Jesper Ørting
- Jesper Lysemose Korsgaard
- Niels Jørgen Strøm
Paper Information
- arXiv ID: 2602.10972v1
- Categories: cs.SE
- Published: February 11, 2026
- PDF: Download PDF