Boards want more throughput for less spend. Many answer with a framework rollout and a heavier cadence of ceremonies. Teams comply, velocity graphs rise, and delivery still slips. The problem is not that standups are bad. It is that rituals cannot compensate for unclear decision rights, unstable priorities, and thin technical enablement. The evidence that exists points to a simple reading: scale by tailoring, not templating.
What the research actually says about scaling
A broad scan of large-scale agile transformations does not give comfort to playbook purists. Dikert, Paasivaara, and Lassenius reviewed 52 publications covering 42 industrial cases and found 35 recurring challenges and 29 success factors (Dikert, Paasivaara and Lassenius, 2016). Almost 90 percent of the included papers were practitioner experience reports rather than academic studies, so the evidence base is uneven. Even so, the pattern is consistent. Success clustered around management support, model customization to local context, training and coaching, and a shift in mindset and alignment. Failure clustered around weak sponsorship, one-size implementation, and lack of buy-in.
Read that list carefully. None of it says a prescribed backlog cadence will save you. It says leaders must adapt methods to the work and the organization they actually have. It also suggests that shipping a framework without coaching, authority, and alignment will produce meetings and dashboards rather than outcomes.
Leadership and priorities, not more meetings
The latest State of DevOps survey reinforces the human and organizational side of performance. The 2024 report associates transformational leadership, defined as leaders who inspire, stimulate intellectually, and support teams, with higher productivity, stronger organizational performance, higher satisfaction, and lower burnout (DORA, 2024). It also reports that instability in organizational priorities is associated with substantial drops in productivity and higher burnout. Crucially, the negative association from unstable priorities persists even when leadership and documentation are rated highly. Translation: you cannot compensate for thrash with better standups or better wikis.
Treat these as correlations, not proof of causation. The DORA data are cross-sectional survey responses. Yet for executives deciding what to fix first, the direction of travel is clear enough. Improve leadership quality and stabilize priorities before you add more process. Protect teams from shifting demands, then ask them to improve their rituals.
Platforms are enablers, not substitutes
Tools are not the answer either, but the right platform can remove friction. The 2024 DORA report links internal developer platforms with better productivity at the individual, team, and organizational level (DORA, 2024). That is an argument for investment in paved roads: common build and deploy tooling, self-service environments, and shared services that shrink cognitive load. It is not a license to skip the hard work of clarifying decision rights or to declare victory because a portal went live. Platforms amplify a good operating model and expose a bad one faster.
Govern by outcomes, tailor the means
If you want teams to customize without drifting into chaos, measure what matters and give them latitude on how. The authors of Accelerate argue for a small set of delivery metrics that capture software performance in practice. Their research program, built on four years of State of DevOps data from 2014 to 2017, claims that specific organizational capabilities drive higher delivery performance and that these outcomes can be measured with a concise set of indicators (Forsgren, Humble and Kim, 2018). The book positions itself explicitly against the view that delivery performance does not confer advantage.
Given the caveats, use these claims as a pragmatic starting point, not revealed truth. Pick a short list of outcome metrics, then let units adapt methods to hit them. For most software groups this means tracking release throughput and lead time, change failure rate, and time to restore. Review these at the level where accountability lives, not as a compliance audit of ceremony checklists.
A practical sequence leaders can run now
Stop mandating rituals first. Start with conditions that make any method worth the effort.
- Stabilize priorities at the portfolio boundary. The DORA finding that instability is associated with persistent drops in productivity and increased burnout should put this at the top of your backlog (DORA, 2024). Institute decision windows and work-in-progress limits for senior stakeholders so that teams can finish what they start.
- Back customization with real sponsorship. The systematic review flags management support and tailoring the model to context as repeated success factors (Dikert, Paasivaara and Lassenius, 2016). Give local leaders the authority to adjust ceremonies, roles, and artifacts within clear guardrails tied to outcomes.
- Invest in coaching and leadership quality. Treat transformational leadership behaviors as a competency to be developed, then measured. The association with better performance and lower burnout is too consistent to ignore, even if it is not causal proof (DORA, 2024).
- Build a platform that removes toil. DORA associates internal developer platforms with productivity gains (DORA, 2024). Fund paved-path tooling that makes the good path the easy path. Do not count the number of services launched. Count the time the platform saves on tasks that used to require tickets and handoffs.
- Govern through outcomes. Use the Accelerate thesis as a governing idea. Pick a small set of delivery metrics, publish them, and let teams adapt. The book’s argument is that capabilities linked to these outcomes matter for advantage. You do not have to accept the causal claim to benefit from the discipline of measuring them (Forsgren, Humble and Kim, 2018).
The temptation in tight markets is to copy a heavyweight framework and call the rollout a transformation. The literature says that is a bet against context and a bet against your own leaders. Scale by subtraction and tailoring instead. Stabilize the work you feed teams, grow leaders who can inspire and support, clear the path with a platform, and measure outcomes rather than ceremony compliance. Do that, and you will have agile in spirit and in results. Fail to do it, and the rituals will calcify while delivery stalls.
Sources
- Dikert, Paasivaara and Lassenius (2016). Challenges and success factors for large-scale agile transformations: A systematic literature review. Finds 35 challenges and 29 success factors across 42 cases, with success tied to management support, customization, coaching, and mindset; cautions that most evidence is practitioner reports. https://www.sciencedirect.com/science/article/pii/S0164121216300826
- DORA / Google Cloud (2024). Accelerate State of DevOps Report 2024. Reports correlations where transformational leadership aligns with higher productivity and lower burnout, while unstable priorities align with lower productivity and higher burnout, and internal developer platforms align with productivity gains. https://dora.dev/research/2024/dora-report/
- Forsgren, Humble and Kim (2018). Accelerate: The Science of Lean Software and DevOps. Argues, based on multi-year State of DevOps data, that specific organizational capabilities drive higher delivery performance and that a small set of outcome metrics can track it; used here as a pragmatic framing with caveats. https://itrevolution.com/product/accelerate/