MBUS 853 — Session 6: Digital Implementation

Block 1 — Session Theme: How You Roll It Out Is the Strategy

From "What's Foundational" to "How Fast Do You Move"

Session 5 asked which architecture decisions are foundational versus tactical. Session 6 asks the implementation question sitting on top of that: once an organization is ready to transform, should it move in one large, company-wide push, or learn its way there through smaller, sequenced experiments? ANZ's CEO Shayne Elliott chose the boldest possible answer — moving the "whole organization to roughly 18 tribes, with 150 squads," a scale one of his own executives compared to running "150 start-ups" simultaneously inside a 180-year-old bank.

Just because a threat is huge doesn't mean a response has to be.

McGrath & McManus's opening thesis — and a direct challenge to the scale and speed of ANZ's own agile rollout.

GE Makes a Third Appearance

The Session 6 article names GE's Predix explicitly as a platform failure — the same case from Session 2 — for a specific implementation reason: Predix "took on way too much too soon, rather than proceeding by finding a good fit for its capabilities and building from there." That's a direct methodological lens to bring into the ANZ discussion: does the case show ANZ making the same too-much-too-soon mistake, just inside its own organizational structure rather than a customer-facing platform?

Block 2 — Article: Discovery-Driven Digital Transformation (McGrath & McManus, HBR 2020)

Learning Your Way to a New Business Model

Built around discovery-driven planning (DDP) — a methodology McGrath developed in the 1990s, later folded into the "lean start-up" toolkit — the article argues traditional firms should reject both extremes: neither a defensive "do nothing" posture nor a start-up-style "bet the farm" pivot (the article opens with Veon's failed 200-person, big-bang mobile platform — 100 staff in Amsterdam, another hundred or so in London). Incumbents have an advantage start-ups don't: the resources to run multiple smaller experiments in parallel, learning faster in aggregate than any single start-up pivot could.

Step 1

Define the operating experience

Before writing a line of code, find what isn't working — workarounds, process stoppages, manual escalations. Best Buy's Renew Blue turned its costliest liability (real estate) into an advantage by asking what a great customer experience required, not what technology was available.

Step 2

Focus on specific problems — outcomes and progress metrics

ROI doesn't reveal customer value. Use a "from-to" table: name the current problem, the target state, and a progress metric — then test and refine assumptions as you go, rather than committing to a fixed ROI projection upfront.

Step 3

Identify your competition — cast a wide net

Industry classification codes are increasingly useless; define competition as an "arena" around the customer job-to-be-done, the way Netflix competes against all leisure time, not just other streaming services.

Step 5

Test your assumptions — failures are lessons too

An assumption checkpoint table reframes "you were wrong, that was a failure" into "was it worth that price to learn what we needed to learn?" — Buffer's two-page landing-page test before building anything is the article's cleanest example.

"Starting big, spending a lot, and assuming you have all the information is likely to produce a full-on attack from corporate antibodies — everything from risk aversion and resentment of your project to simple resistance to change."

— McGrath & McManus — a close description of the "frozen middle" resistance ANZ's own executives named directly

The size paradox: the article's own logic (incumbents can run many small parallel experiments better than a single start-up pivot) is exactly what ANZ's "150 start-ups" framing gestures at — the open question for discussion is whether ANZ actually ran 150 independent, learn-as-you-go experiments, or one very large synchronized rollout wearing 150 different badges.

Block 3 — Case: Australia and New Zealand Banking Group: The Agile Transformation

From Toxic Culture to "The ANZ Way"

ANZ traces to 1835 (Bank of Australasia); by 2017 it was one of Australia's "Big Four" banks, with over 50,000 employees, A$897B in total assets, and operations in 34 countries. Shayne Elliott became CEO in January 2016 after Mike Smith's nine-year tenure (2007–2015) — an era marked by an unmet "super-regional" Asia ambition, a 10x increase in Asia-based headcount that didn't deliver, and, starting September 2015, a series of scandals: trading-number manipulation, drug and alcohol use in dealing rooms under lax oversight, and an ASIC lawsuit over "unconscionable conduct and market manipulation" of the bank bill swap reference rate, eventually settled for over $50 million.

The ANZ Way and NWOW

In July 2017, Elliott sent a 3,500-word letter to all 50,000 employees launching "The ANZ Way" — three priorities (purpose-driven change, digital technology for trust and experience, and a simpler/better-capitalized/better-balanced bank) underpinned by new cultural values: ICARE (Integrity, Collaboration, Accountability, Respect, Excellence). The organizational vehicle for this was New Ways of Working (NWOW), ANZ's term for its agile transformation, reorganizing the bank into tribes (divisions), squads (multidisciplinary, autonomous, ~10-person teams with end-to-end ownership), and a shift from functional managers to coaches and product owners chosen for adaptability rather than seniority.

Scale and Speed of the Rollout

Milestone	Detail
2016	Katherine Bray uses agile "sprints" to deliver Apple Pay to customers in just 10 weeks — the proof point that earns her the lead role in the transformation that follows (predates NWOW's announcement by roughly a year)
May 2017	NWOW announced
Early 2018 (planned)	Agile-based transformation extended to the entire organization — not a pilot, a full rollout target
FY2018 target	Scale from 3,000 to 13,000 additional people in agile squads and tribes
FY2018 actual	Over 9,000 Australian people in technology divisions working in agile teams — strong traction in tech specifically, but concentrated there rather than achieved bank-wide, with New Zealand still only "planned to extend"
May 2019	Elliott pauses the rapid expansion to "improve the existing expansion stage" — the case's own decision point

Christian Venter, one of the executives who led the project, described the ambition directly: "We're moving the whole organization to roughly 18 tribes, with 150 squads, and I'd like to think of that as we've got 150 start-ups running." Physical space was redesigned to match — a "squad living room" replaced the executive floor, with billboards, miniature office-furniture models, and a pop-up cinema with beanbags.

Block 4 — Challenges to the Transformation

Where NWOW Ran Into Friction

The "Frozen Middle"

Maile Carnegie, ANZ's executive of digital banking, named this directly: middle managers with outdated skillsets, no longer sufficient for the new environment, who resisted change — a distinct obstacle from culture itself, and one the case treats as a structural, not just attitudinal, problem.

Job Insecurity and Morale

Employees were told existing roles might no longer exist, even as management framed the restructuring as an "exciting chapter" — a messaging gap that upset staff and reportedly led some to pause major life decisions like home financing.

Funding Model Mismatch

Squads were reorganized around continuous, iterative delivery, but funding remained yearly or project-based — a structural mismatch the case frames as Venter's own view: the collision between new ways of working and old ways of financing. ANZ's fix-in-progress: capacity-based funding tied to squad headcount.

Uneven Applicability

NWOW simply did not work in some areas, such as contact centres, branches, and dealing rooms — the case's own account of what Elliott found. The case frames the real challenge as recognizing where agile applies, not assuming it scales uniformly across every function.

The Financial Backdrop

-18%

ANZ's cash profit decline, driven by $1.077B in restructuring costs

$6.41B

FY2016/17 profit, up 12% year-over-year — the strong year before restructuring costs hit

-5%

half-year profit decline, FY2018/19 vs. same period prior year

Elliott's own words capture the tension the case leaves open: "Although we had a strong business, the external environment was changing faster than we were and our customers, the community, and our shareholders expected much more from us." ANZ's managing director of products, Katherine Bray, offered the counter-frame: "There is of course a risk that day one — everything's done, high five in the corridor, move on, transformation over. Well that's not the nature of this... we will never be done."

Block 5 — Memo Protocol: Team Case Study Memo #4

Format Reminder Before the Team Writes

Due Thursday 11:59pm before Session 6, based only on the ANZ case, two pages, 11-point font, written wholly by the team.

To Shayne Elliott, CEO — the case's own closing questions are framed around his decision to pause and reassess the transformation's pace, making him the clean single decision maker.

Issues Exactly 5, each grounded in a specific case fact — the frozen-middle resistance, the funding-model mismatch, NWOW's uneven applicability across functions, the 18% cash-profit decline, and the toxic-culture legacy are all strong, distinct candidates.

Problem/Decision 40–60 words on the underlying cause — consider whether the root issue is rollout speed/scale, funding structure, or unresolved cultural trust, since these pull toward different recommendations.

Alternatives Exactly 3, mutually exclusive, feasible, not simultaneous, not status quo.

Criteria Exactly 3 standards for judging the alternatives.

Evaluation/Recommendation 120–140 words, pros/cons per alternative per criterion, no table, ending in a justified pick.

Actions Exactly 3 steps not already taken in the case.

Case-only constraint: this case (published 2020, covering events through May 2019) predates ANZ's actual subsequent history — it's easy to accidentally reach for real-world knowledge of what ANZ did after 2019. Stay strictly inside the case's own timeline and its own closing, unresolved questions.

Academic integrity — GenAI is banned in submitted work for this course. The diagnostic analysis below is discussion prep, not memo text — the team's actual submission must be written independently.

Block 6 — Case Diagnostic: Issues, Decision, Position (Discussion Prep)

Applying the Case Prep Protocol

Step 1 — Who and What

Decision maker: Shayne Elliott, CEO of ANZ. Core challenge: in May 2019, Elliott paused the rapid expansion of an agile transformation he'd championed for two years — needing to decide whether NWOW's next phase should scale faster, scale more selectively, or fundamentally change how it's funded and led, at a moment when both culture and financial performance remain unresolved.

Step 2 — Candidate Issues Grounded in Case Facts

Rollout speed outran organizational readiness. The target of 13,000 additional people in agile teams by end of 2018, and a plan to extend to "the entire organization" in early 2018, moved faster than the frozen-middle resistance and funding-model problems could be resolved.
The frozen middle was named but not yet solved. Carnegie's own diagnosis — outdated skillsets among functional managers — describes a structural talent gap, not a motivational one, and the case doesn't show a concrete plan to close it.
Funding structure lagged the operating model. Squads were designed for continuous iteration; funding remained annual/project-based until Venter's capacity-based funding fix, itself still in progress as of the case's close.
NWOW's applicability was uneven, and this was discovered late. The case's account that Elliott found agile did not work in contact centres, branches, and dealing rooms suggests those functions weren't tested first before broader commitments were made.
The financial case for the transformation isn't yet proven. An 18% cash-profit decline (driven by $1.077B in restructuring costs) and a 5% half-year profit decline complicate the argument that NWOW is delivering value fast enough to justify its pace and cost.

Step 3 — A Position

Underlying problem, one sentence: ANZ designed a genuinely values-led, well-intentioned transformation (The ANZ Way, ICARE, NWOW) but rolled it out at big-bang scale and speed — company-wide targets, a full-organization 2018 deadline — before running the kind of small, sequenced, discovery-driven tests McGrath and McManus recommend, so problems that a pilot would have surfaced early (funding mismatch, uneven functional fit, frozen-middle resistance) instead surfaced at full scale, after most of the capital and organizational disruption had already been spent.

Counterargument to weigh: One could argue a slower, more incremental rollout was never realistic given ANZ's starting point — a bank recovering from public trading scandals and an ASIC lawsuit needed a decisive, visible, company-wide signal of cultural change, not a quiet pilot in one back-office function that regulators, customers, and the public would barely notice. The strongest response has to weigh whether the scandal-recovery urgency genuinely required scale-first sequencing, or whether it only required visible commitment (the ANZ Way letter, ICARE) while the operational rollout itself could still have been sequenced more incrementally.

Counterargument to weigh — autonomy vs. risk control: NWOW's core design principle is to push authority down and out, to tribes and squads, rather than keep it centralized with divisional managers — arguably the opposite instinct of what a bank recovering from its own scandal should reach for. The scandals that triggered The ANZ Way in the first place involved traders who, per the case, "continued to enjoy virtual autonomy and lucrative bonuses" under lax oversight; NWOW then deliberately expands autonomy further into the organization. The case itself flags this tension without resolving it, noting that "as authority was delegated to tribes and scrums rather than centralized to divisional managers, ANZ had to ensure that risks were still under control." On this reading, the real open question isn't rollout speed at all — it's whether decentralizing authority was ever the right instinct for an institution whose recent crisis was itself a story of insufficiently supervised autonomy.

Step 4 — 30-Second Cold-Call Answer

ANZ's transformation ran into trouble not because The ANZ Way's vision was wrong, but because Shayne Elliott's team rolled out New Ways of Working at full organizational scale — a company-wide, early-2018 deadline and a 13,000-person target — before testing whether agile actually fit every function. The case's own evidence shows this: technology absorbed agile fastest, reaching over 9,000 people, while Elliott himself later found NWOW did not work in contact centres, branches, and dealing rooms — a mismatch a smaller pilot would likely have caught before ANZ took an 18% hit to cash profit on $1.077 billion in restructuring costs. On top of that, funding stayed annual and project-based while squads were built for continuous delivery, and the "frozen middle" Maile Carnegie named directly never got a concrete retraining plan. So the May 2019 pause isn't really a pause on agile itself — it's ANZ discovering, at full scale and real financial cost, exactly the kind of assumption failures McGrath and McManus's discovery-driven approach is designed to surface cheaply and early.

Block 7 — Discussion Questions & Sharp Answers

Likely Professor Questions

Framing to expect: (1) Was ANZ's agile transformation the right solution to its external and internal challenges? (2) How could a 180-year-old, deeply rooted culture be successfully transformed? (3) Should Elliott scale NWOW further, or fix its foundations first?

Q1: Was ANZ's "150 start-ups" framing consistent with a genuinely discovery-driven approach, or was it a big-bang rollout in disguise?

More big-bang than discovery-driven, on the article's own terms. A true DDP approach tests assumptions with from-to tables and assumption checkpoints before scaling — ANZ instead set a company-wide numeric target (13,000 additional people by end of 2018) and a full-organization deadline (early 2018) up front, then discovered mid-rollout that agile didn't work in several functions. Note too that Apple Pay — the "proof of concept" behind Bray's promotion — was delivered in 2016, a full year before NWOW was even announced (May 2017), so it wasn't itself a DDP-style pilot of the agile operating model; it was evidence used to justify going big afterward. The "150 start-ups" language borrows the vocabulary of incremental experimentation without the sequencing discipline the article actually prescribes.

Redamo Labs' IAM rollout succeeded partly because each new capability was piloted with a small group before company-wide deployment — the opposite sequencing from ANZ's simultaneous, organization-wide squad rollout.

Q2: Carnegie names the "frozen middle" as a key obstacle. Is this a culture problem or a talent problem?

Primarily a talent problem wearing a culture-problem's clothes. Carnegie's own words are specific: outdated skillsets, "reaching the point of being no longer sufficient" — a capability gap, not simply resistance to change. Treating it purely as a culture problem (more communication, more ICARE messaging) would under-invest in the actual fix these managers need: retraining, redeployment, or replacement with people who have the adaptive skills the coach/product-owner roles require. ANZ's own hiring shift (managers chosen for adaptability over seniority) is an implicit admission of this.

At Prodigy Education, some long-tenured managers needed structured upskilling (not just encouragement) to lead cross-functional, data-driven teams effectively — treating it as a mindset issue alone would have missed the real skills gap.

Q3: Should Elliott resume rapid NWOW expansion after the May 2019 pause, or use the pause to fundamentally rethink the funding and applicability model first?

Fix foundations before resuming expansion. The case's own evidence shows NWOW succeeded fastest and most convincingly in the function it fit best (technology — over 9,000 people, the strongest traction of any function) and struggled or failed outright in functions poorly suited to it (contact centres, branches, dealing rooms). Resuming expansion at the prior pace would repeat that same mismatch at greater scale, while the funding model (capacity-based funding, still being rolled out) and frozen-middle talent gap remain unresolved — precisely the kind of foundational fixes a genuine pause is supposed to enable.

This is the strongest question to lead class discussion with — it's a direct, falsifiable read of the case's own data (technology's outperformance vs. other functions' explicit failure) rather than a values-based opinion.

Block 8 — Participation Hooks & Taju's Edge

How to Contribute Distinctively

Open Strong

Don't open with "ANZ's culture needed to change." Open with the scale mismatch: ANZ set a company-wide, full-organization agile deadline before it had tested whether agile even worked outside technology — the article calls this exact pattern out by name, using GE's Predix as the cautionary example.

Push the Consensus

Class will likely say "the frozen middle resisted change." Push further: Carnegie's own words describe a skills gap, not a motivation gap — the fix ANZ needed was retraining and redeployment, not more change-management messaging.

Bridge to Session 2

The article's direct citation of GE's Predix ("took on way too much too soon") is a gift — explicitly connect ANZ's company-wide rollout target back to GE Digital's own overreach from Session 2, using the article's own words as the bridge.

Taju's Edge — Redamo Labs

Cutting verification time from 8 to 2 minutes came from piloting changes with a small group first, then scaling — a direct, lived counter-example to ANZ's simultaneous, company-wide squad rollout.

Taju's Edge — Prodigy Education

A/B testing culture across a 150M-user platform is itself a discovery-driven methodology in practice — testing assumptions cheaply before committing resources at scale, exactly what McGrath and McManus prescribe.

Taju's Edge — Stutern

Growing from zero to 100,000+ users at 40% MoM growth required constant reassessment of what was working before scaling further — a bootstrapped version of the same "learn before you scale" discipline ANZ needed at 50,000-employee scale.

Block 9 — Reflections Journal Prep (Fill In After Class)

Concept Must Come From Live Slides

As with Sessions 2–5, the journal's concept half needs to be captured live from Prof. Mufti's Session 6 slide deck. The example half is ready to pair with whatever concept the session covers.

Concept — Fill In From Session 6 Slides

[3–7 word identification — write after class]

150–200 words, from class slides only. Implementation-focused candidates to listen for: agile methodology, discovery-driven planning, pilot-then-scale sequencing, or change-management pacing frameworks — write down whichever the professor actually presents.

Candidate Example — Ready to Pair With Whatever Concept Fits

Redamo Labs — Piloting Before Committing the Whole Team

When a new fraud-detection workflow was proposed for the IAM platform, the instinct was to roll it out to the full 12-person team at once, given how promising the early results looked. Instead, it launched with two team members for three weeks first. That small pilot surfaced a specific edge case — the workflow flagged a category of legitimate high-volume accounts as suspicious — that would have created a much larger backlog and eroded trust in the tool if it had shipped to the whole team immediately. Fixing it before the wider rollout meant the eventual full-team adoption went smoothly, with none of the "we don't trust this tool" skepticism that a rockier launch would have created. The instinct to move fast at full scale is usually the wrong instinct — the fix is almost always cheaper to find in a small pilot than in a company-wide rollout.

Block 10 — Key Takeaways

What to Walk Away Knowing

A huge threat doesn't require a huge, all-at-once response. McGrath and McManus's opening thesis directly challenges the instinct behind ANZ's company-wide, fixed-deadline rollout.

Incumbents' real advantage is running many small experiments in parallel, not one big bet. ANZ's "150 start-ups" language borrowed this idea's vocabulary without fully adopting its sequencing discipline.

Resistance to change is sometimes a skills gap wearing a culture-problem's clothes. The "frozen middle" needed retraining and redeployment, which is a different fix than more change-management communication.

Operating model changes need matching changes to funding and metrics — not just structure. ANZ's agile squads couldn't function well under annual/project-based funding, exactly the "from-to" mismatch the article's own methodology is designed to catch early.

Looking Ahead — Session 7: Digital Renewal

→ Session 7 (Renewal)

Session 6 asks "how should a transformation be sequenced?" Session 7 (How Incumbents Survive and Thrive; Washington Post case) asks "how does a legacy incumbent sustain reinvention over decades, not just one transformation cycle?"

↔ Recurring Thread: Prove, Then Scale

Harley's York-first sequencing (Session 5) and Pernod Ricard's TLO periods (Session 4) both anticipated this session's discovery-driven logic — ANZ is the sharpest case yet of what happens when that discipline is skipped.

Memo #5 Due Before Session 7

The team's fifth and final Case Study Memo (Washington Post) is due the Thursday before Session 7 at 11:59pm.