In Brief
- Declaring a program agile and running an agile program are not the same thing — and the difference is visible in behaviour, not vocabulary.
- Genuine agile practice produces a 42% project success rate against waterfall's 13%, and a failure rate of 11% against 59% — a structural gap, not a marginal one.
- Sidky's values-to-practices model is directionally correct but incomplete: the relationship between values and practices runs in both directions, not one.
- Psychological safety, self-organisation, and respect for people are not cultural aspirations — they are the operating conditions without which the performance gap cannot be realised.
- Executives who want genuinely adaptive teams are asking a governance question — what authority has been placed where the information actually lives.
Ahmed Sidky’s agile mindset model (2014) is clean, logical, and directionally correct. It is also widely taught. The model describes a continuum from mindset to values, values to principles, principles to practices, practices to frameworks such as Scrum, Kanban and Scaled Agile Framework (SAFe). The arrow runs left to right. It teaches well and accepts easily.
The model is incomplete, and the incompleteness matters. A linear sequence implies causation: get the values right at the start and the practices will follow. That logic is exactly what allows a program to declare itself agile, install a framework, rename the roles and the meetings, and continue operating exactly as before, all while genuinely believing the transition is underway. Practitioners sometimes call this pattern “wagile” — waterfall execution dressed in agile vocabulary. Zen Ex Machina’s diagnostic work begins with this distinction: declaring a program agile and running one are not the same thing — and the difference is visible in behaviour, not vocabulary.
What the research actually shows
The case for genuine agile practice is well documented.
The Standish Group’s CHAOS studies have tracked the outcomes of more than 50,000 technology projects across a quarter-century. In the 2020 dataset, agile projects achieved a 42% success rate against waterfall’s 13%. Agile projects failed outright at 11% against waterfall’s 59% (The Standish Group, 2020). The gap widens as project size increases; medium and large programs show the most dramatic divergence. A 29-point success gap and a 48-point failure gap are structural differences, not statistical noise.
McKinsey’s 2021 global research across more than 800 transforming organisations found that fully agile organisations — those with both structural and cultural transformation in place — achieved a five-to-tenfold increase in the speed of decision-making and change, a 10–30% improvement in customer satisfaction and operational efficiency, and a 20 to 30 percentage point increase in employee engagement (Aghina, Handscomb, Salo & Thaker, 2021). McKinsey’s caveat is the more important finding. Partial transformations, where agile practices are adopted without the accompanying operating model and cultural conditions, do not produce these outcomes. They produce something closer to the industry average, sometimes worse.
A 42% success rate versus 13%. An 11% failure rate versus 59%. The performance gap between genuine agile practice and waterfall is structural, not marginal.
The mechanism that converts governance structure into delivery outcome is decision latency: the time between a decision becoming necessary and the decision being made. The Standish Group’s 2018 CHAOS report identified decision latency as the single largest determinant of project success across its dataset, outweighing methodology choice. Organisations with short decision latency achieved success rates substantially higher than those with poor decision latency, regardless of whether they called their approach agile or not. That finding reorders the conversation: methodology is secondary, governance conditions are primary.
The distinction the research surfaces, and that practitioners rarely make clearly, is this. The performance gap is not between agile and waterfall. It sits between genuine agile practice and everything else, including agile theatre performed with commitment.
Why installing frameworks is not enough
A more complete model maps two hemispheres that must function together rather than a left-to-right flow from mindset to practices. The left hemisphere is logical structure: demonstrating agility through empiricism, managing the system of work rather than the people, optimising the flow of value, building alignment and synchronisation across teams, holding customer centricity at the centre of decisions. Beneath those actions sit the frameworks that give them operational form: Scrum, Kanban, Lean, XP, Nexus, Scrum@Scale, LeSS, SAFe. The right hemisphere is social-emotional. It contains the four values of the Agile Manifesto (Beck et al., 2001) and the twelve principles that follow from them, and between values and principles a layer Sidky’s model omits entirely: ethics, psychological safety, self-organisation, and respect for people.
The relationship between the two hemispheres runs in both directions. Festinger’s (1957) formulation of cognitive dissonance established that the relationship between held values and observable behaviour is bidirectional. People change their stated values to match their behaviour almost as readily as they change their behaviour to match their values. In organisational contexts the consequence is direct: a team operating under waterfall governance, with fixed scope, fixed budget, and escalation chains, will internalise the values that waterfall produces. The agile values on the wall will not survive contact with the operating model.
The patterns are consistent: teams that find Scrum too constraining strip it to vocabulary and call the result agile. Teams that want flexibility claim a pragmatic approach without the track record to know what to be pragmatic about. The common thread is that the declaration of agility replaces the examination of whether the conditions for agility are in place. These are not character failures in the individuals expressing them. They are the predicted output of organisations running the left hemisphere with the right hemisphere stripped out.
Eloranta, Koskimies and Mikkonen (2016), in their empirical study of Scrum anti-patterns, found that the most harmful deviations are the ones that preserve the observable form of Scrum events while removing the authority and conditions those events depend on. A Sprint Review that cannot change direction. A Sprint Retrospective whose improvement actions require approval that never arrives. The events remain on the calendar. The decision-making they were designed to enable does not.
The declaration replaces the work
West, Gilpin, Grant and Anderson (2011) named the dominant pattern at Forrester Research as water-Scrum-fall: agile practices layered over waterfall governance, producing neither the predictability of waterfall nor the responsiveness that agile is designed to create. Dikert, Paasivaara and Lassenius (2016) reviewed 52 publications covering 42 industrial cases of large-scale agile transformation. Their finding: reverting to old structural patterns while maintaining surface adoption was the most consistently reported failure mode, including in well-resourced transformations led by experienced practitioners.
“I think I’m agile; therefore, I am”.
– NOT Descartes
Cogito ergo sum. Believing you are agile is not the same as operating as one.
The psychological mechanism that makes this pattern so durable is one Samantha Boardman, a psychiatrist and Clinical Instructor at Weill-Cornell Medical College, identified in her analysis of positive thinking. A program that believes it is agile is less likely to examine the concrete actions those practices require or the obstacles preventing them. The declaration of agility becomes a substitute for the work of becoming agile (Boardman, 2016).
A program that believes it is agile is less motivated to examine whether the conditions for agility are actually in place. The declaration of agility functions as a substitute for the work of becoming agile — and research on the psychology of positive thinking offers a clear mechanism for why: organisations that have conceptually achieved their transformation goals become less likely to identify the concrete obstacles preventing them (Boardman, 2016).
Alami and Krancher (2022) studied 39 Scrum practitioners across multiple organisations and published their findings in Empirical Software Engineering. The social antecedents most reliably associated with quality outcomes were collaboration, psychological safety, accountability, and transparency. These are exactly the conditions that fail to materialise when the right hemisphere is absent. Their specific failure conditions follow the same pattern: inconsistent implementations, cultural constraints, and inaccessibility of end-users. In each case the left hemisphere is in place. The right hemisphere is not.
A program that believes it is agile is less likely to examine the concrete actions required and the obstacles preventing them. The declaration becomes the substitute.
What adaptive teams actually require
The organisations that close the performance gap — the ones producing the 42% success rates rather than the 13% — are not the ones with the most mature frameworks or the most comprehensive training programs. They are the ones in which both hemispheres are operating. The distinguishing measure is decision latency.
The Standish Group’s 2018 CHAOS report, drawing on more than 50,000 project profiles, identified decision latency as the single largest factor in project outcomes. It outweighed methodology choice in every project size category the report examined. A project that generates one decision for every $1,000 in labour cost, with a pipeline that requires escalation through three or four management layers, accumulates enormous overhead. The overhead appears nowhere in the plan; it shows up as slower delivery, more rework, and teams waiting rather than working (The Standish Group, 2018).
Moe, Dingsøyr and Dybå (2010), in a nine-month field study of a Scrum team published in Information and Software Technology, found that delegating decision authority to the operational level increases the speed of addressing problems and adapting to changing conditions. The mechanism is self-organisation, which the right hemisphere of the ZXM model names explicitly and which Sidky’s linear sequence omits. Self-organisation is also the layer that agile theatre most reliably destroys. When an organisation installs Scrum frameworks but routes every substantive decision upward, it has produced the left hemisphere with the right hemisphere missing. The teams become faster at executing tasks. Their capacity to adapt remains unchanged.
Decision latency is the cost of absent self-organisation. It appears nowhere in the plan and everywhere in the result.
What makes the difference between agile theatre and genuine agile practice? Governance structure — specifically, whether decision-making authority sits where the information lives. An executive who genuinely wants adaptive teams — teams that respond to new information rather than defending an existing plan — is not asking a methodology question. They are asking whether their governance model has placed decision authority where the information actually lives.
Adaptive teams require decision-making authority to sit where the information lives — in the team, not a governance layer above it. That is not a framework question. It is a question about what authority has been placed where, and by whom.
These are governance decisions, and they sit with the executive rather than the delivery teams. Frameworks and practices — the left hemisphere — function as designed when the right hemisphere conditions are in place. Without those conditions, what the executive gets is water-Scrum-fall with better vocabulary: agile theatre performed with commitment, producing outcomes indistinguishable from the waterfall it replaced. Zen Ex Machina’s governance advisory work begins with these structural questions — not with framework selection.
Governance decisions already made are either making adaptability structurally possible or preventing it — and that determination has nothing to do with which framework the delivery teams are running.
References
Aghina, W., Handscomb, C., Salo, O., & Thaker, S. (2021, May 25). The impact of agility: How to shape your organization to compete. McKinsey & Company.
Alami, A., & Krancher, O. (2022). How Scrum adds value to achieving software quality? Empirical Software Engineering, 27(165).
Beck, K., Beedle, M., van Bennekum, A., Cockburn, A., Cunningham, W., Fowler, M., Grenning, J., Highsmith, J., Hunt, A., Jeffries, R., Kern, J., Marick, B., Martin, R. C., Mellor, S., Schwaber, K., Sutherland, J., & Thomas, D. (2001). Manifesto for agile software development.
Boardman, S. (2016, November 8). Positive thinking? Overrated. Psychology Today.
Dikert, K., Paasivaara, M., & Lassenius, C. (2016). Challenges and success factors for large-scale agile transformations: A systematic literature review. Journal of Systems and Software, 119, 87–108.
Eloranta, V.-P., Koskimies, K., & Mikkonen, T. (2016). Exploring ScrumBut — An empirical study of Scrum anti-patterns. Information and Software Technology, 74, 194–203.
Festinger, L. (1957). A theory of cognitive dissonance. Row, Peterson and Company. (Reissued 1962, Stanford University Press)
Moe, N. B., Dingsøyr, T., & Dybå, T. (2010). A teamwork model for understanding an agile team: A case study of a Scrum project. Information and Software Technology, 52(5), 480–491.
Sidky, A. (2014). The agile mindset [conference presentation]. Agile conference.
The Standish Group. (2018). CHAOS report: Decision latency theory — it is all about the interval. The Standish Group.
The Standish Group. (2020). CHAOS report: Beyond infinity. The Standish Group.
West, D., Gilpin, M., Grant, T., & Anderson, A. (2011). Water-Scrum-fall is the reality of agile for most organizations today [technical report]. Forrester Research.
