Trend AnalysisManagement & Business
Agile at Scale: Why Enterprise Agile Transformations Succeed, Struggle, or Stall
Agile methodologies were designed for small, co-located software teams. Applying them to large organizations with hundreds of developers, complex product portfolios, and established governance structu...
By Sean K.S. Shin
This blog summarizes research trends based on published paper abstracts. Specific numbers or findings may contain inaccuracies. For scholarly rigor, always consult the original papers cited in each post.
Agile methodologies were designed for small, co-located software teams. Applying them to large organizations with hundreds of developers, complex product portfolios, and established governance structures is a fundamentally different challengeโone that has generated an entire industry of frameworks (SAFe, LeSS, Nexus, Spotify Model) and a mixed record of results.
Hutter, Brendgens, and Gauster (2023) provide one of the most detailed longitudinal case studies of enterprise agile transformation, following a multinational corporation through its multi-year journey. Their analysis identifies persistent challenges including cultural resistance, organizational inertia, and the complexity of coordinating multiple teams across enterprise-scale agile transformations. The most critical finding is that the scaling phase is where most transformations failโnot because agile practices do not work at team level but because the organizational structures surrounding the teams (budgeting cycles, governance processes, incentive systems, middle management roles) were designed for waterfall development and actively resist agile principles. Successful transformations required changing these surrounding structures simultaneously with team practices, a scope of organizational change that most leadership teams underestimate.
Azonuche and Enyejo (2024) evaluate the impact of agile scaling frameworks specifically in large-scale fintech software development, where rapid delivery, regulatory compliance, and system reliability must coexist. Their review finds that scaled agile frameworks improve development speed and team satisfaction but show more ambiguous effects on code quality and system reliability. The productivity gains come primarily from reduced handoff delays and faster feedback loops, while quality concerns arise from the pressure to deliver frequently without proportional investment in testing infrastructure and technical debt management. The study identifies a common anti-pattern: organizations that adopt agile velocity metrics without corresponding quality gates end up shipping faster but accumulating technical debt that eventually slows them down more than the original waterfall process.
Stan, ศรฎศu, and Iamandii (2025) synthesize the challenges and solutions literature, identifying five persistent obstacles to enterprise agile adoption: cultural resistance (particularly from middle management whose roles are most disrupted), coordination complexity (synchronizing dozens of teams without creating bureaucratic overhead), technical dependencies (monolithic architectures that prevent independent team delivery), compliance requirements (regulatory frameworks designed for plan-driven development), and measurement difficulties (traditional project metrics do not capture agile value delivery). For each obstacle, the authors catalog solutions that have shown effectiveness, noting that no single solution addresses all challenges and that the appropriate approach depends on organizational context, industry, and the specific agile framework adopted.
The synthesis across these studies challenges both agile evangelism and agile skepticism. Enterprise agile transformation can deliver genuine benefitsโfaster time to market, better responsiveness to change, higher employee engagementโbut it requires organizational change far deeper than adopting stand-up meetings and sprint reviews. The organizations that succeed are those that treat agile not as a project management methodology but as an organizational design philosophy, willing to restructure governance, budgeting, and management practices around the same principles they are asking development teams to adopt.
๋ฉด์ฑ
์กฐํญ: ์ด ํฌ์คํธ๋ ์ ๋ณด ์ ๊ณต ๋ชฉ์ ์ ์ฐ๊ตฌ ๋ํฅ ๊ฐ์์ด๋ค. ํน์ ์ฐ๊ตฌ ๊ฒฐ๊ณผ, ํต๊ณ, ์ฃผ์ฅ์ ํ์ ์ ์๋ฌผ์์ ์ธ์ฉํ๊ธฐ ์ ์ ์๋ฌธ ๋
ผ๋ฌธ์ ํตํด ๋ฐ๋์ ๊ฒ์ฆํด์ผ ํ๋ค.
์ ์์ผ(Agile) ๋ฐฉ๋ฒ๋ก ์ ์๊ท๋ชจ์ ๋์ผ ์ฅ์ ๊ทผ๋ฌด ์ํํธ์จ์ด ํ์ ์ํด ์ค๊ณ๋์๋ค. ์๋ฐฑ ๋ช
์ ๊ฐ๋ฐ์, ๋ณต์กํ ์ ํ ํฌํธํด๋ฆฌ์ค, ํ๋ฆฝ๋ ๊ฑฐ๋ฒ๋์ค ๊ตฌ์กฐ๋ฅผ ๊ฐ์ง ๋๊ท๋ชจ ์กฐ์ง์ ์ด๋ฅผ ์ ์ฉํ๋ ๊ฒ์ ๊ทผ๋ณธ์ ์ผ๋ก ๋ค๋ฅธ ๋์ ๊ณผ์ ์ด๋ฉฐ, ์ด๋ก ์ธํด ๋ค์ํ ํ๋ ์์ํฌ(SAFe, LeSS, Nexus, Spotify Model)๋ก ์ด๋ฃจ์ด์ง ํ๋์ ์ฐ์
์ ์ฒด๊ฐ ํ์ฑ๋์๊ณ ๊ฒฐ๊ณผ๋ ์๊ฐ๋ฆฌ๊ฒ ๋ํ๋ฌ๋ค.
Hutter, Brendgens, Gauster(2023)๋ ๋ค๊ตญ์ ๊ธฐ์
์ ์๋
์ ๊ฑธ์น ์ฌ์ ์ ์ถ์ ํ๋ฉฐ ๊ธฐ์
์์ค์ ์ ์์ผ ์ ํ์ ๊ดํ ๊ฐ์ฅ ์์ธํ ์ข
๋จ ์ฌ๋ก ์ฐ๊ตฌ ์ค ํ๋๋ฅผ ์ ๊ณตํ๋ค. ์ด๋ค์ ๋ถ์์ ๋ฌธํ์ ์ ํญ, ์กฐ์ง์ ๊ด์ฑ, ๊ทธ๋ฆฌ๊ณ ๊ธฐ์
๊ท๋ชจ์ ์ ์์ผ ์ ํ ๊ณผ์ ์์ ์ฌ๋ฌ ํ์ ์กฐ์จํ๋ ๋ณต์ก์ฑ ๋ฑ ์ง์์ ์ธ ๋์ ๊ณผ์ ๋ค์ ์๋ณํ๋ค. ๊ฐ์ฅ ํต์ฌ์ ์ธ ๋ฐ๊ฒฌ์ ๋๋ถ๋ถ์ ์ ํ์ด ์ค์ผ์ผ๋ง(scaling) ๋จ๊ณ์์ ์คํจํ๋ค๋ ๊ฒ์ธ๋ฐ, ์ด๋ ์ ์์ผ ์ค์ฒ์ด ํ ์์ค์์ ์๋ํ์ง ์๊ธฐ ๋๋ฌธ์ด ์๋๋ผ, ํ์ ๋๋ฌ์ผ ์กฐ์ง ๊ตฌ์กฐ(์์ฐ ํธ์ฑ ์ฃผ๊ธฐ, ๊ฑฐ๋ฒ๋์ค ํ๋ก์ธ์ค, ์ธ์ผํฐ๋ธ ์์คํ
, ์ค๊ฐ ๊ด๋ฆฌ์ ์ญํ )๊ฐ ํญํฌ์(waterfall) ๋ฐฉ์์ ๊ฐ๋ฐ์ ์ํด ์ค๊ณ๋์ด ์ ์์ผ ์์น์ ๋ฅ๋์ ์ผ๋ก ์ ํญํ๊ธฐ ๋๋ฌธ์ด๋ค. ์ฑ๊ณต์ ์ธ ์ ํ์ ์ํด์๋ ํ ์ค์ฒ ๋ฐฉ์๊ณผ ์ด๋ฌํ ์ฃผ๋ณ ๊ตฌ์กฐ๋ฅผ ๋์์ ๋ณํ์์ผ์ผ ํ๋๋ฐ, ์ด๋ ๋๋ถ๋ถ์ ๋ฆฌ๋์ญ ํ์ด ๊ณผ์ํ๊ฐํ๋ ์กฐ์ง ๋ณํ์ ๋ฒ์์ด๋ค.
Azonuche์ Enyejo(2024)๋ ์ ์ํ ๋ฐฐํฌ, ๊ท์ ์ค์, ์์คํ
์ ๋ขฐ์ฑ์ด ๊ณต์กดํด์ผ ํ๋ ๋๊ท๋ชจ ํํ
ํฌ(fintech) ์ํํธ์จ์ด ๊ฐ๋ฐ์์ ์ ์์ผ ์ค์ผ์ผ๋ง ํ๋ ์์ํฌ์ ์ํฅ์ ํ๊ฐํ๋ค. ์ด๋ค์ ๊ฒํ ์ ๋ฐ๋ฅด๋ฉด ์ค์ผ์ผ๋ ์ ์์ผ(scaled agile) ํ๋ ์์ํฌ๋ ๊ฐ๋ฐ ์๋์ ํ ๋ง์กฑ๋๋ฅผ ํฅ์์ํค์ง๋ง, ์ฝ๋ ํ์ง๊ณผ ์์คํ
์ ๋ขฐ์ฑ์๋ ๋ ๋ชจํธํ ์ํฅ์ ๋ฏธ์น๋ค. ์์ฐ์ฑ ํฅ์์ ์ฃผ๋ก ์ธ๊ณ(handoff) ์ง์ฐ ๊ฐ์์ ๋น ๋ฅธ ํผ๋๋ฐฑ ๋ฃจํ์์ ๋น๋กฏ๋๋ ๋ฐ๋ฉด, ํ์ง ์ฐ๋ ค๋ ํ
์คํธ ์ธํ๋ผ ๋ฐ ๊ธฐ์ ๋ถ์ฑ(technical debt) ๊ด๋ฆฌ์ ๋ํ ๋น๋ก์ ํฌ์ ์์ด ์ฆ์ ๋ฐฐํฌ๋ฅผ ํด์ผ ํ๋ ์๋ฐ์์ ๋ฐ์ํ๋ค. ์ด ์ฐ๊ตฌ๋ ํํ ์ํฐ ํจํด(anti-pattern)์ ์๋ณํ๋๋ฐ, ์ด๋ ํด๋น ํ์ง ๊ฒ์ดํธ(quality gate) ์์ด ์ ์์ผ ์๋(velocity) ์งํ๋ฅผ ์ฑํํ ์กฐ์ง์ด ๊ฒฐ๊ตญ ๋ ๋น ๋ฅด๊ฒ ์ถ์ํ์ง๋ง ๊ธฐ์ ๋ถ์ฑ๋ฅผ ์ถ์ ํ์ฌ ๊ฒฐ๊ตญ ๊ธฐ์กด์ ํญํฌ์ ๋ฐฉ์๋ณด๋ค ๋ ๋๋ ค์ง๊ฒ ๋๋ค๋ ๊ฒ์ด๋ค.
Stan, ศรฎศu, Iamandii(2025)๋ ๋์ ๊ณผ์ ๋ฐ ํด๊ฒฐ์ฑ
๋ฌธํ์ ์ข
ํฉํ์ฌ ๊ธฐ์
์ ์ ์์ผ ๋์
์ ๊ฐ๋ก๋ง๋ ๋ค์ฏ ๊ฐ์ง ์ง์์ ์ธ ์ฅ์ ๋ฌผ์ ์๋ณํ๋ค: ๋ฌธํ์ ์ ํญ(ํนํ ์ญํ ์ด ๊ฐ์ฅ ํฌ๊ฒ ๋ณํํ๋ ์ค๊ฐ ๊ด๋ฆฌ์๋ก๋ถํฐ), ์กฐ์ ๋ณต์ก์ฑ(๊ด๋ฃ์ ์ค๋ฒํค๋๋ฅผ ์์ฑํ์ง ์๊ณ ์์ญ ๊ฐ์ ํ์ ๋๊ธฐํํ๋ ๊ฒ), ๊ธฐ์ ์ ์์กด์ฑ(๋
๋ฆฝ์ ์ธ ํ ๋ฐฐํฌ๋ฅผ ๋ฐฉํดํ๋ ๋ชจ๋๋ฆฌ์(monolithic) ์ํคํ
์ฒ), ์ปดํ๋ผ์ด์ธ์ค(compliance) ์๊ฑด(๊ณํ ์ฃผ๋ ๊ฐ๋ฐ์ ์ํด ์ค๊ณ๋ ๊ท์ ํ๋ ์์ํฌ), ๊ทธ๋ฆฌ๊ณ ์ธก์ ์ ์ด๋ ค์(์ ํต์ ์ธ ํ๋ก์ ํธ ์งํ๊ฐ ์ ์์ผ์ ๊ฐ์น ์ ๋ฌ์ ํฌ์ฐฉํ์ง ๋ชปํจ)์ด๋ค. ๊ฐ ์ฅ์ ๋ฌผ์ ๋ํด ์ ์๋ค์ ํจ๊ณผ๊ฐ ์
์ฆ๋ ํด๊ฒฐ์ฑ
๋ค์ ๋ชฉ๋กํํ๋ฉด์, ๋จ์ผ ํด๊ฒฐ์ฑ
์ด ๋ชจ๋ ๋์ ๊ณผ์ ๋ฅผ ํด๊ฒฐํ์ง๋ ๋ชปํ๋ฉฐ ์ ์ ํ ์ ๊ทผ ๋ฐฉ์์ ์กฐ์ง์ ๋งฅ๋ฝ, ์ฐ์
, ๊ทธ๋ฆฌ๊ณ ์ฑํํ ํน์ ์ ์์ผ ํ๋ ์์ํฌ์ ๋ฐ๋ผ ๋ฌ๋ผ์ง๋ค๊ณ ์ธ๊ธํ๋ค.
์ด๋ฌํ ์ฐ๊ตฌ๋ค์ ์ข
ํฉ์ ์ ์์ผ ์ง์ง๋ก ๊ณผ ์ ์์ผ ํ์๋ก ๋ชจ๋์ ์๋ฌธ์ ์ ๊ธฐํ๋ค. ๊ธฐ์
์ ์ ์์ผ ์ ํ์ ์์ฅ ์ถ์ ์๊ฐ ๋จ์ถ, ๋ณํ์ ๋ํ ๋ ๋์ ๋์๋ ฅ, ๋์ ์ง์ ์ฐธ์ฌ๋ ๋ฑ ์ง์ ํ ์ด์ ์ ์ ๊ณตํ ์ ์์ง๋ง, ์ด๋ ์คํ ๋์
(stand-up) ๋ฏธํ
๊ณผ ์คํ๋ฆฐํธ(sprint) ๋ฆฌ๋ทฐ๋ฅผ ๋์
ํ๋ ๊ฒ์ ํจ์ฌ ๋์ด์๋ ์ฌ์ธต์ ์ธ ์กฐ์ง ๋ณํ๋ฅผ ํ์๋ก ํ๋ค. ์ฑ๊ณตํ๋ ์กฐ์ง์ ์ ์์ผ์ ํ๋ก์ ํธ ๊ด๋ฆฌ ๋ฐฉ๋ฒ๋ก ์ด ์๋ ์กฐ์ง ์ค๊ณ ์ฒ ํ์ผ๋ก ๋ฐ์๋ค์ด๊ณ , ๊ฐ๋ฐ ํ์๊ฒ ์๊ตฌํ๋ ๊ฒ๊ณผ ๋์ผํ ์์น์ ์ค์ฌ์ผ๋ก ๊ฑฐ๋ฒ๋์ค, ์์ฐ ํธ์ฑ, ๊ด๋ฆฌ ์ค์ฒ ๋ฐฉ์์ ๊ธฐ๊บผ์ด ์ฌ๊ตฌ์ฑํ๋ ์กฐ์ง์ด๋ค.
References (3)
[1] Hutter, K., Brendgens, F.-M. & Gauster, S. (2023). Scaling organizational agility: keyย insights from an incumbent firm's agile transformation. Management Decision, 61, md-05-2022-0650.
[2] Azonuche, T.I. & Enyejo, J.O. (2024). Evaluating the Impact of Agile Scaling Frameworks on Productivity and Quality in Large-Scale Fintech Software Development. International Journal of Scientific Research and Modern Technology, 3(6), 449.
[3] Stan, N.-M., ศรฎศu, A. & Iamandii, A.-F. (2025). Agile Project Management in Large Organizations: Challenges and Solutions. Proceedings of the International Conference on Business Excellence, 2025, 0172.