Scrum rituals are on the calendar. Progress, oddly, is not. Boards want speed, product leaders fear another slip, and teams feel the heat. The reflex is to add more ceremonies, new templates and stricter process policing. The evidence points elsewhere: measure outcomes, manage the system of work, and fix the few variables that repeatedly show up in the data.
Measure outcomes, not rituals
A decade of survey research synthesized in Accelerate argues that a small set of software delivery performance metrics can be tracked reliably, and that specific organizational capabilities are linked statistically to better results (Forsgren, Humble and Kim, 2018). The authors analyzed multi‑year State of DevOps data with statistical methods and make a clear claim: these capabilities are not just associated with performance, they causally drive it. That is a strong statement and comes from a book that summarizes survey‑based evidence. Treat it as a hypothesis you can test inside your own system rather than a universal law.
If you are still counting story points and ceremony attendance, you are measuring the wrong thing. The point is not more agile, it is faster and more stable flow of work. Move governance to a short list of outcome metrics, then ask what constrains those outcomes in your context. When metrics go up without adding headcount, keep the change. When they do not, stop, learn and try a different intervention.
Two variables most firms neglect
The 2024 State of DevOps report is blunt about one of the biggest drags on performance: unstable organizational priorities. Respondents who reported frequent shifts in priorities also reported substantially lower productivity and higher burnout, and that association persisted even when leadership quality and documentation were rated highly (DORA, 2024). The study is cross‑sectional and correlational, not causal. Even with that caveat, the signal is hard to ignore. If your portfolio thrashes, no ritual will rescue throughput.
Leadership style also shows up strongly. The same report links transformational leadership, leaders who inspire, challenge, and support teams, with higher productivity, stronger organizational performance, greater satisfaction, and lower burnout (DORA, 2024). Again, this is association, not proof of causation. Yet it is consistent with the idea that agile is a management discipline before it is a method. If executives and line managers do not create clarity, remove obstacles, and protect focus, the backlog will move slowly no matter how many retros you run.
There is a complementary systems lever worth calling out. Internal developer platforms are associated with higher productivity at the individual, team, and organizational levels in the same dataset (DORA, 2024). Tooling is not a silver bullet, and this finding is correlational. It does suggest that investing in shared services that reduce cognitive load and waiting time is more likely to pay off than inventing a new ceremony.
Why large transformations stall
The largest agile programs often look impressive on paper. They also fail quietly. A systematic review of 52 publications covering 42 industrial cases of large‑scale agile transformation found that nearly 90 percent of the evidence base consisted of practitioner reports rather than academic studies, a reminder that scaling recipes rest on thin science (Dikert, Paasivaara and Lassenius, 2016). The review catalogued 35 distinct challenges across nine categories and 29 success factors across 11 categories. Despite the heterogeneity, several themes recur: sustained management support, customizing the model to context, meaningful training and coaching, and a shift in mindset and alignment rather than mere process rollout (Dikert, Paasivaara and Lassenius, 2016).
Put simply, most failures are not caused by teams running the wrong standup. They stem from leaders underestimating the organizational rewiring needed to achieve flow. Copy‑paste frameworks struggle because they assume ceremony fidelity produces outcomes. The review suggests the opposite. Fit the model to your constraints, invest in the human and managerial system, and align on goals you can actually hold steady for long enough to matter.
A practical operating cadence
If you want agile beyond the ritual, change what you manage each week.
- Replace ceremony audits with a simple delivery outcomes dashboard. Use the small set of delivery performance measures popularized by Accelerate as your north star, and treat the book’s causal claims as working hypotheses to validate in your environment (Forsgren, Humble and Kim, 2018).
- Stabilize intake. Make one executive decision that matters: priorities hold for a defined window. The 2024 State of DevOps report associates unstable priorities with lower productivity and higher burnout even in otherwise strong contexts, which implies that stability is a first‑order constraint to manage (DORA, 2024).
- Train leaders, not only teams. Invest in the behaviors captured under transformational leadership. The report links them to better productivity, stronger organizational performance and lower burnout. Correlational evidence is still useful when you are deciding where to spend scarce coaching dollars (DORA, 2024).
- Remove friction with platforms and shared services. The same report links internal developer platforms with higher productivity at multiple levels. Instrument carefully and watch stability as you change tooling, then keep what moves outcomes in your context (DORA, 2024).
- Customize consciously. The large‑scale review warns against treating framework compliance as success. Build from management support, context‑specific tailoring, and real coaching, then adapt as your metrics and teams’ experience warrant (Dikert, Paasivaara and Lassenius, 2016).
This is a different posture from traditional transformation. You are not trying to win at ceremonies. You are running a portfolio of operating hypotheses about what will improve flow, testing them against outcomes you can see, and making leadership accountable for reducing thrash and protecting focus. That model is closer to how modern product teams work. It is also how you prevent agile from degenerating into performance art.
The payoff is not abstract. When priorities stop shifting mid‑sprint, burnout falls and productivity improves in the patterns reported across thousands of respondents, even if the underlying relationships are correlational rather than causal (DORA, 2024). When leaders model the behaviors teams need, the same patterns show up. When you instrument delivery outcomes and tailor the way of working instead of installing a template, you align with what the most comprehensive review of large‑scale transformations actually observed in practice (Dikert, Paasivaara and Lassenius, 2016). Agile earns its keep when leaders manage the conditions for flow and measure what changes as a result. Everything else belongs in the calendar, not the board pack.
Sources
- Dikert, Paasivaara and Lassenius (2016). Challenges and success factors for large-scale agile transformations: A systematic literature review. Synthesizes 52 publications across 42 cases, finds a weak evidence base and highlights management support, customization, coaching, mindset and alignment as recurring success factors. https://www.sciencedirect.com/science/article/pii/S0164121216300826
- DORA / Google Cloud (2024). Accelerate State of DevOps Report 2024. Cross-sectional survey linking unstable priorities with lower productivity and higher burnout, and transformational leadership and internal developer platforms with better outcomes. https://dora.dev/research/2024/dora-report/
- Forsgren, Humble and Kim (2018). Accelerate: The Science of Lean Software and DevOps. Summarizes multi-year State of DevOps research, argues a small set of delivery performance metrics and specific capabilities causally drive higher software delivery performance. https://itrevolution.com/product/accelerate/