Model the Transformation You Expect Employees to Deliver
Antonio Nieto-Rodriguez · article · 2026-07-22 source ↗ Reader 🖥️
Organizational transformations succeed or fail not because employees resist change but because senior leaders either do or do not visibly perform, in their own calendars and decisions, the same risk-taking, uncertainty-tolerance, and willingness-to-reverse-course they are asking employees to adopt.
practical · argumentative essay with embedded case studies — A consulting-derived prescriptive essay: it diagnoses a cause of transformation failure (leaders not modeling the behavior they demand), supports the diagnosis with cited benchmarking research and two contrastive case studies, then converts the diagnosis into a diagnostic instrument (the three tests) the reader can apply to their own leadership team. Practical throughout — every element serves the imperative in the title.
Structure
| Part | Role |
|---|---|
| Opening vignette: the CEO's all-hands rhetoric about becoming agile/digital/customer-obsessed evaporates within a week as the senior team reverts to its old operating rhythm. | motivating scene / problem statement |
| Debunking 'employees resist change' using McKinsey (5.3x success multiplier when leaders visibly model behavior, only 43% do) and Prosci (executive sponsorship the single biggest success factor since 1998; 79% vs 27% success by sponsor quality) benchmarking data, relocating the bottleneck from the workforce to the top. | evidentiary premise establishing the causal claim |
| Framing via the author's 'project economy' thesis (from his book Powered by Projects): the project-driven organization is replacing the operational machine, which raises the stakes on leaders operating differently. | theoretical/contextual scaffolding |
| Diagnosis of the mechanism: legacy operating forums (QBRs, capital approval committees, annual planning, performance management) reward exactly the behaviors the transformation is supposed to replace, creating a contradiction employees register immediately. | causal mechanism |
| Positive case: Piyush Gupta at DBS Bank (2009-2024) — personally visiting branches incognito, making experimentation a leadership KPI, running an anonymous 'Tell Piyush' feedback channel, entering an internal AI competition and finishing without special treatment. | narrative / supporting case |
| Additional shorter positive anecdotes: a European energy CEO chairing the riskiest diversification bet, a pharma CFO reassigning her strongest controller to an unproven initiative, a retail bank CEO staying in the room through mixed early results. | generalizing corroboration |
| Diagnosis of failure mode: modeling failure is rarely deliberate sabotage but an accumulation of individually 'prudent' decisions (still requiring a full business case for a 6-week experiment, handing the riskiest initiative to the safest leader, withholding best people from the transformation). | causal mechanism, negative case |
| Negative case: António Horta-Osório's appointment as Credit Suisse chairman (2021) to fix a culture of unaccountability, his personal breach of the same accountability standard (Covid rules, private jet use), resignation within nine months, and Credit Suisse's collapse into UBS's takeover fifteen months later. | narrative / contrastive case |
| The diagnostic instrument: three tests (calendar test, uncertainty test, cancellation test) that operationalize 'is the senior team inside the transformation or watching from outside.' | prescriptive core / conclusion |
| Closing case: an unnamed industrial company stalled for two years until a new CEO personally took ownership of the two hardest initiatives, after which stuck decisions cleared within weeks. | narrative / closing corroboration |
| Closing restatement: 'leaders cannot delegate a transformation' — you can sponsor, fund, communicate, govern it, but without visibly living the change the organization will stop pretending to. | conclusion |
Key terms
- Modeling the transformation
A leader's own calendar time, decisions, and personal risk exposure visibly matching the behaviors (experimentation, risk-taking, course-correction) demanded of employees — distinguished sharply from communicating, funding, or governing the change. - The calendar test
A specific diagnostic: what share of the senior team's actual monthly time (not stated priority) goes to the transformation versus running the existing business, and whether the CEO's own time is on the riskiest bet or only the safe ones. - The uncertainty test
Whether the senior team has any shared, lived experience of operating together inside a genuinely unresolved bet, as opposed to reviewing and governing the transformation from outside. - The cancellation test
Whether the senior team has publicly killed an initiative it previously championed, with lessons stated out loud — used as the most visible proof that changing course costs nothing when evidence changes. - Project-driven organization / project economy
The author's broader framework (from Powered by Projects) describing an organizational form where value creation is increasingly project work ('change the business') rather than repeatable operational work ('run the business'), which is displacing the operational machine as the dominant structure. - Active and visible executive sponsorship
Cited from Prosci's benchmarking research as the single biggest contributor to change success across 12 studies since 1998 — quantified, not rhetorical (79% vs 27% success rate by sponsor effectiveness).
Propositions
- explicit"Employees resist change" is a comforting explanation leaders offer for transformation failure, but it is not the actual cause.conf: high
- explicitWhen senior leaders visibly model the behaviors they ask of employees, transformations are 5.3 times more likely to succeed, yet only 43% of leaders actually do so (McKinsey research on 1,000+ transformations).conf: high
- explicitActive and visible executive sponsorship has been the single biggest contributor to change success in every Prosci benchmarking edition since 1998; projects with an extremely effective sponsor meet or exceed objectives 79% of the time versus 27% with an extremely ineffective sponsor, and absent sponsorship is the most-cited obstacle to success.conf: highsupportsDORA Community Discussion - Operationalizing DORA within a large organization · p5 — This article's Prosci statistic (executive sponsorship the single biggest contributor to change success since 1998) corroborates, at benchmarking-research scale, the DORA case's concrete finding that a 12,000-person change effort depended critically on securing an effective executive sponsor (a new CIO) as part of its coalition.
- explicitThe workforce is not, by and large, the bottleneck in failed transformations; the bottleneck is at the top.conf: high
- explicitLegacy operating forums (quarterly business reviews, capital approval committees, annual planning cycles, performance management systems) structurally punish the exact behaviors (risk-taking, fast experimentation, changing course) that a transformation requires.conf: high
- explicitFailure to model a transformation is typically not deliberate sabotage but the cumulative effect of individually defensible, prudent-looking decisions (requiring a business case for a small experiment, staffing the riskiest bet with the safest leader, withholding top talent from the transformation).conf: high
- explicitThe CEO personally sponsoring the most ambiguous, highest-uncertainty initiative in the portfolio (rather than the safest one) is a distinguishing feature of the most successful project-driven organizations the author works with.conf: medium
- explicitPiyush Gupta's personal, visible participation in DBS Bank's transformation (incognito branch visits, treating experimentation as a leadership KPI, the 'Tell Piyush' channel, entering the DeepRacer AI competition as a non-expert) was what made the transformation strategy 'stick,' beyond the strategy's content itself.conf: highsupportsOrdinary Engineers, Not Heroic Inventors · p14 — Both claims converge on the same mechanism: oreilly-ordinary-engineers holds that effective enterprise AI transformation needs leadership that 'gets its own hands dirty'; this article's DBS case (Gupta visiting branches incognito, entering the DeepRacer competition himself) is a concrete, non-AI-specific instance of the same hands-on-leadership mechanism driving transformation success.
- explicitAntónio Horta-Osório's personal breach of the accountability standard he was hired to instill at Credit Suisse (Covid quarantine violations, private jet scrutiny) is presented as the reason the cultural transformation he led never recovered momentum, contributing to the bank's collapse into UBS's takeover.conf: medium
- explicitA leadership team's calendar allocation to transformation work typically stays under 10% on first measurement even when rhetoric implies it should be closer to half.conf: medium
- explicitA leadership team cannot understand what it is asking employees to do under uncertainty unless it has itself sat together inside a live, unresolved bet — this cannot be delegated or simulated in a workshop.conf: high
- explicitPublicly cancelling a previously championed initiative, with lessons stated aloud, is the most visible act of organizational learning a leadership team can perform; healthy portfolios retire two or three bets a year while unhealthy ones keep everything alive indefinitely.conf: high
- explicitThe three tests (calendar, uncertainty, cancellation) all reduce to one underlying question: whether the senior team is inside the transformation or watching it from outside.conf: high
- explicitA transformation cannot be delegated: it can be sponsored, funded, communicated, and governed, but if leaders are not visibly living the change, the organization will eventually stop pretending to as well.conf: high
- implicitcorroboratedEmployees calibrate their own commitment to a transformation by observing leadership behavior rather than leadership rhetoric, and register the gap between the two quickly.conf: mediumsupports ←Frontline Workers Know How to Solve Your Organization's Biggest Problems · p7 — Both find that employees calibrate trust and commitment by observing whether leaders visibly act, not by what leaders say they will do.
Arguments
The primary determinant of transformation success is whether senior leaders visibly model the demanded behaviors themselves, not whether employees are willing to change. strength: medium
Explicit premises:
- McKinsey: visible leadership modeling makes transformations 5.3x more likely to succeed; only 43% of leaders model.
- Prosci: executive sponsorship quality has been the largest single contributor to change success since 1998, with a 79%-vs-27% success gap by sponsor effectiveness.
- The most-cited obstacle to transformation success is absent sponsorship, not workforce resistance.
Implicit premises:
- Correlational benchmarking data (leaders who model transformations succeed more often) can be read as evidence of a causal mechanism (modeling causes success) rather than a byproduct of some third factor (e.g., leaders confident enough to model behavior may also be better strategists or operate in less troubled organizations).
- 'Visible modeling' and 'effective sponsorship' as measured by McKinsey and Prosci are close enough proxies for the same underlying behavior that the two studies can be combined into one argument.
Inferences:
- If modeling behavior is both necessary and the largest measured lever, then interventions aimed at employee mindset or communication are lower-leverage than interventions aimed at leadership behavior.
- Therefore diagnosing and fixing leadership behavior should precede or dominate diagnosing and fixing employee resistance.
Credit Suisse's cultural transformation failed because Horta-Osório failed to personally embody the accountability standard he was hired to instill. strength: medium
Explicit premises:
- Horta-Osório was recruited specifically to fix 'a cultural problem that translated into a lack of accountability' and stated the new direction was personal responsibility and accountability.
- He personally breached Covid quarantine rules twice and drew scrutiny over private-jet use — conduct directly contradicting the standard he was hired to model.
- He resigned nine months into the role, and the cultural transformation 'never recovered momentum.'
- Credit Suisse ceased to exist as an independent institution fifteen months after his resignation.
Implicit premises:
- The chairman's personal conduct, rather than the underlying governance and risk-control failures (Greensill, Archegos) that predated his appointment, is the operative cause of the transformation's failure to recover.
- No credible successor or interim structure could substitute for the chairman's personal credibility once it was breached.
- The 15-month gap between his resignation and the UBS takeover is short enough, and causally connected enough, to count as a consequence of the failed cultural transformation rather than an independent outcome of the pre-existing Greensill/Archegos losses and market confidence collapse.
Inferences:
- If a symbolic personal breach can derail an already-launched cultural transformation, then leader credibility is a single point of failure the organization has limited ability to hedge against.
- Therefore boards should treat personal conduct scrutiny of transformation sponsors as equivalent in importance to their strategic competence.
Narratives
- Piyush Gupta's 15-year transformation of DBS Bank (from a bank customers nicknamed 'Damn, Bloody, Slow' to one of Innosight's ten most transformative organizations of the 2010s), achieved by personally visiting branches incognito as a customer, making experimentation a leadership KPI (1,000 experiments run in one year with most leadership participating), running a quarterly anonymous 'Tell Piyush' suggestion channel (about a third of suggestions leading to policy changes), and entering the bank's internal AWS DeepRacer AI competition himself, finishing in the top 100 among DBS's own employees.
→ Serves as the article's positive proof case: strategy content (digitize, mandate experimentation, build startup culture) was 'familiar in outline,' so the narrative's work is to isolate personal, visible leadership modeling as the differentiating variable that made the strategy 'stick.' - António Horta-Osório was recruited as Credit Suisse's chairman in April 2021 specifically to repair a culture of unaccountability following the Greensill and Archegos losses (the latter costing about $5.5 billion), publicly committing to 'a culture of personal responsibility and accountability,' then resigned nine months later after an internal investigation found he had personally breached Covid quarantine rules twice and drawn scrutiny over private-jet use; the cultural transformation never recovered, and Credit Suisse was forced into a UBS takeover fifteen months after his resignation.
→ Serves as the article's negative/contrastive proof case, structurally paired with DBS: it implicitly supports the proposition that a leader's personal conduct — not strategy quality, board mandate, or stated commitment — is what determines whether a transformation an organization badly needs actually lands. - An unnamed large industrial company spent nearly two years 'in transformation' with a sound strategy and an active program office but little to show for it, because the executive team reviewed the effort monthly rather than leading it; a new CEO then took personal ownership of the two hardest initiatives, moved them to the top of his own calendar, and worked in the sessions where the real trade-offs were made, after which the rest of the team followed within weeks and previously stuck decisions cleared.
→ Closing corroboration that converts the article's diagnostic claim into a before/after natural experiment within a single organization, implicitly supporting the proposition that modeling behavior (not new strategy or new resourcing) was the variable that changed between the stalled period and the unstuck period. - Three brief, unnamed vignettes: a European energy company CEO personally chairing the group's most uncertain diversification bet while letting safer programs run without him; a global pharmaceutical CFO moving her strongest controller onto an unproven AI initiative; a retail bank CEO joining the working sessions of a digital venture of uncertain survival and staying through mixed early results.
→ Generalizes the DBS pattern beyond one famous, extensively studied case, implicitly supporting the claim that visible personal modeling is a replicable leadership behavior observable across industries and company sizes, not an idiosyncratic feature of Gupta or DBS specifically.
Questions addressed
- ✅ Why do transformations actually fail, if not because of employee resistance?
Because senior leaders do not visibly perform, in their own time and decisions, the behaviors (risk-taking, tolerance for uncertainty, willingness to cancel failed bets) they demand of employees, while continuing to govern through legacy forums that reward the opposite behaviors. - ✅ How can a leadership team tell, concretely, whether it is actually modeling the transformation rather than merely endorsing it?
Apply the three tests: the calendar test (time allocation and whose bet the CEO owns), the uncertainty test (shared lived experience of an unresolved bet), and the cancellation test (whether anything previously championed has been publicly killed). - ⭕ Is Credit Suisse's collapse best explained by governance/risk failures (Greensill, Archegos) or by a failure of leadership modeling?
The article names both but does not adjudicate their relative weight; it asserts the modeling failure is 'at its heart' without ruling out that the antecedent governance and risk failures were independently sufficient to doom the bank regardless of Horta-Osório's personal conduct. - ⭕ What should a leadership team do once it has failed the three tests — is there a remediation process, or only the diagnostic?
Not addressed; the article ends on the diagnostic ('fail them, and you've already told your people...') without describing how a team moves from failing to passing, beyond the implicit suggestion (via the industrial-company case) that a single leader taking personal ownership of the hardest initiatives can unstick things quickly. - ⭕ Does the prescription scale below the CEO/chairman level — can a business-unit leader or middle manager 'model the transformation' with the same effect if the leaders above them are not doing so?
Not addressed at all; every example in the article is a CEO or chairman-level actor.
Gaps
- incompleteThe article never specifies how a leadership team gets from failing the three tests to passing them, beyond one anecdote (a new CEO personally taking ownership of two hard initiatives). It gives no account of what happens when a team recognizes it is failing but the CEO who would need to change their calendar is the same person unwilling or unable to do so — the article's own logic (leaders cannot delegate this) implies there is no remediation path that does not run through the top person's own willingness, but this implication is never stated or examined.
Read next: Read against literature on board-level accountability and CEO succession/removal for cultural failure (relevant given the Credit Suisse case itself ended in the chairman's own removal), and against change-management frameworks (e.g., Kotter, ADKAR) that specify sequencing and remediation steps rather than only diagnostics. - incompleteEvery case in the article — Gupta, Horta-Osório, the unnamed industrial-company CEO, the energy-company CEO, the pharma CFO, the retail-bank CEO — sits at CEO, chairman, or C-suite level. The article's own opening claim ('the bottleneck is at the top') is never tested against a case where the bottleneck is one level below the top: a business-unit leader or VP whose own boss is not modeling the transformation. It is left unaddressed whether visible modeling by a sub-CEO leader can still move their own team's behavior, or whether the effect requires the very top of the hierarchy to model it.
Read next: Read alongside middle-management- or informal-leadership-focused accounts of change agency operating without top-level air cover, to test whether 'modeling' works bottom-up/mid-up or only top-down. - incompleteThe McKinsey and Prosci statistics (5.3x more likely to succeed; 79% vs 27%) are presented as if the causal arrow runs cleanly from leader modeling to transformation success. The article does not consider or rule out reverse causation (transformations that are already succeeding for other structural reasons — better resourcing, less internal politics, a healthier starting culture — may simply free leaders to model behavior visibly, rather than the modeling causing the success) or confounding (leaders confident and secure enough to visibly take personal risk may be systematically different, and more effective for unrelated reasons, from leaders who are not).
Read next: Read alongside the original McKinsey and Prosci methodology write-ups to check whether either study attempts to control for confounds or establish direction of causality, or is purely correlational self-reported benchmarking. - uninformedThe article treats the Credit Suisse collapse as if the modeling failure were the decisive cause ('at its heart is Horta-Osório's failure to embody and model the transformation'), while its own text names the Greensill and Archegos losses (the latter alone costing about $5.5 billion) as having already 'laid bare' the cultural problem before Horta-Osório's appointment. It does not engage with the possibility that Credit Suisse's fate was already largely sealed by those prior losses and the resulting collapse in market and counterparty confidence, independent of whether the chairman personally modeled accountability afterward.
Read next: Read banking-sector post-mortems on Credit Suisse's collapse (regulatory reports, the Swiss FINMA assessment) that weigh governance/risk-control failures against leadership/cultural factors, to establish whether the modeling-failure framing is proportionate to its actual causal weight.