The Discipline of Calculated Risk

How Retail’s Best Leaders Structure Their Most Consequential Bets — and What Separates Disciplined Boldness from Reckless Conviction

-

-

“The leaders who failed were not reckless. They were disciplined — about the wrong things. The discipline itself was the problem.”

On the morning of March 28, 2007, Circuit City store managers across the United States assembled their teams and read from a corporate script. In a single coordinated action, 3,400 of the company’s most experienced sales associates were escorted from their stores. They had done nothing wrong except earn too much money.

Eighteen months later, Circuit City filed for Chapter 11 bankruptcy. A company that Jim Collins had once featured in Good to Great — the NYSE’s top-performing stock of the 1980s — had made a catastrophic decision. Not through recklessness. Through the rigorous, board-approved, analytically sound application of a discipline pointed entirely in the wrong direction.

That is the argument this article makes, and it is more unsettling than a story about impulsive leadership: the retail executives who destroyed their companies in the last two decades were, in most cases, disciplined people. They had risk frameworks. They ran scenario analyses. They hired consultants. They failed not because they lacked process, but because their process was built to price one type of risk — the risk of acting — while systematically ignoring another: the compounding, invisible, ultimately fatal risk of not acting.

The discipline of calculated risk is something different. It is a structured capability for making the right bets, at the right scale, with the right architecture — and for pricing what it costs to refuse them. This article maps that discipline through ten retail cases, distills it into a working framework, and asks the question every leader in a disrupted category must eventually confront: do you have the discipline, or do you only have the process?

THE CRITICAL DISTINCTION

Calculated Risk vs. Reckless Risk: Drawing the Line

The most common misreading of this article’s argument is the simplest: that the lesson is to take bigger risks. It is not. Ron Johnson’s JCPenney overhaul was bold, committed, and publicly named. It was also reckless — and the distinction between that and Hubert Joly’s Best Buy transformation is structural, not a matter of outcome. Both leaders were confident. Both were experienced. Only one had the discipline.

CALCULATED RISK	RECKLESS RISK
Hypothesis: A specific, written belief about what will change and why.	Hypothesis: A conviction held without evidence — or evidence from the wrong reference group.
Test: A pilot-sized test to answer the question before full-scale — reversible if wrong.	Test: None. Full national rollout from Day One with no reversibility architecture.
Horizon: Inaction priced across the full strategic time horizon, not just the current cycle.	Horizon: Short-term metrics only. The long-term cost of the decision is never modeled.
Owner: A named leader who has staked personal credibility on the outcome.	Owner: Diffuse accountability — or a single leader insulated from the consequences.
Example: Joly's price-matching pilot. Lacik's lab-grown diamond test. Horowitz's five-year silent rebuild.	Example: Johnson’s JCPenney overhaul. Victoria’s Secret’s years of half-measures without named commitment.

The framework this article proposes that “Diagnose, Design, Decide — is not an argument for boldness. It is an argument for structured boldness: bets designed to be tested before they become irreversible, sized to be survived if wrong, and named to create the institutional commitment that prevents post-decision drift.

WHERE THE DISCIPLINE FAILED

Five Cases That Show How Good Leaders Make Fatal Decisions

What follows is not a gallery of reckless executives. Every organization here had experienced leadership, analytical capability, and access to the same competitive intelligence that their more successful peers were reading. These are failures of discipline, specifically, the failure to apply the right discipline at the moment the decision became irreversible.

1.  Circuit City: When the Wrong Number Gets All the Attention

In 2007, Circuit City fired 3,400 of its most experienced salespeople — not for poor performance, but because they earned above market rate. The logic looked unambiguous on a spreadsheet: save $775 million over seven years. What the spreadsheet did not capture was that Circuit City’s entire differentiation over Best Buy rested on the expertise of those very people. That expertise was not a cost item. It was the product. The same holiday season, same-store sales fell twelve percent.

FAILURE 01  ·   CIRCUIT CITY   ·  The Expertise Liquidation, 2007–2009

The Diagnose discipline requires pricing both sides of the ledger. Circuit City’s analysis was rigorous on costs and entirely absent on competitive impact. No model asked what would happen to conversion and retention when differentiation walked out the door simultaneously across 640 stores. Best Buy, facing the identical environment, invested in exactly the expertise Circuit City eliminated.

OUTCOME:  Chapter 11 filed November 2008. All 567 stores closed, 34,000 jobs eliminated. Best Buy became the category’s sole survivor.

2.  Blockbuster: The Right Analysis, the Wrong Time Horizon

In 2000, Reed Hastings offered Blockbuster a partnership: Netflix would run its online business for $50 million. Blockbuster’s late-fee revenue stood at $800 million annually — real, visible, quarterly. CEO John Antioco chose the certain. His analysis was not wrong. It was conducted across the wrong time horizon. A ten-year model would have shown that digital distribution would eliminate the late-fee model entirely. Not reduce it. Eliminate it.

FAILURE 02  ·   BLOCKBUSTER   ·  The $50 Million Refusal, 2000–2010

The board commissioned a rigorous analysis of the cost of action. No equivalent analysis was commissioned for the cost of inaction across a ten-year horizon. Blockbuster Online launched in 2004 — four years after the optimal window had closed. By then, Netflix had the subscription habit, the brand trust, and four years of data Blockbuster would never recover.

OUTCOME:  Chapter 11 filed September 2010. Netflix’s market capitalisation reached $300 billion. The $50 million declined in 2000 remains the most consequential single decision in retail history.

3.  Sears: Fifteen Years of Quarterly Decisions That Added Up to Nothing

Sears’ failure is not the story of a bold bet gone wrong. It is the story of the complete absence of any bold bet at all. For fifteen years, successive management teams consistently chose the path that preserved short-term earnings while quietly destroying the conditions for long-term survival. Every decision was financially defensible in the quarter it was made. Not one addressed the question the business actually needed to answer: in a world where Amazon, Walmart, and Target were investing billions, what was the reason for Sears to exist?

FAILURE 03  ·   SEARS HOLDINGS   ·  Fifteen Years of Optimisation Becoming Liquidation, 2003–2018

The Diagnose stage was never conducted at board level because doing so would have required confronting an answer the board was unwilling to act on. Every quarter, the cost of transformation grew larger and the capital available to fund it grew smaller. The risk that destroyed Sears was not the risk it took. It was the compounding risk it refused — quarter by quarter, across fifteen years of individually rational decisions.

OUTCOME:  Chapter 11 filed October 2018, approximately $11 billion in liabilities. Market capitalisation fell from a $30 billion peak to near zero.

4.  Victoria’s Secret: The Brand That Refused to See Itself

By 2016, every signal available to Victoria’s Secret’s leadership pointed in the same direction. The body-positivity movement was not a fringe trend — it had reshaped the category. Aerie, Savage X Fenty, and a generation of direct-to-consumer challengers were capturing precisely the consumers Victoria’s Secret was losing. Internal research confirmed the brand was reading as exclusionary to an expanding majority of its own customer base. The diagnosis was available. The decision to act on it was not made. CEO Jan Singer and Chief Marketing Officer Ed Razek chose, publicly and repeatedly, to defend the existing brand identity rather than reframe it. The 2018 Fashion Show — the brand’s annual set piece — drew its lowest ratings ever. Razek’s comment that transgender models did not belong in the show became the definitive public marker of a brand that had confused heritage with strategy.

FAILURE 04  ·   VICTORIA’S SECRET  ·  The Commitment That Never Came, 2016–2023

The Decide discipline requires naming the bet and making it irrevocable. Victoria’s Secret diagnosed the problem — the evidence was unambiguous by 2016 — but never named a committed response. Leadership made minor tactical adjustments (introducing extended sizing, quietly retiring the Fantasy Bra format) without staking the brand on a defined new identity. The repositioning that eventually arrived under Martin Waters in 2021 — VS Collective, new brand ambassadors, revised marketing language — came five years after the diagnosis and two years after the delistings and store closures that had already done structural damage. The Decide failure is not the absence of action. It is the absence of named, irrevocable commitment. Half-measures taken without a public hypothesis do not constitute the Decide discipline — they are its most expensive substitute.

OUTCOME:  U.S. market share fell from 31.7% in 2013 to approximately 20% by 2021. The 2019 Fashion Show was cancelled entirely. L Brands spun off Victoria’s Secret in 2021 as a standalone public company at a fraction of its former valuation. Savage X Fenty, founded in 2018, reached a $1 billion valuation in three years. The brand that had defined the category for two decades ceded its position not to a single disruptor but to a generation of smaller brands that committed to the identity Victoria’s Secret refused to name.

5.  JCPenney: A Coherent Vision, Deployed Without a Test

Ron Johnson arrived at JCPenney in 2011 with genuine credibility — he had built the Apple Store. His vision was coherent: transparent everyday pricing, brand boutiques, a customer experience that respected the shopper’s intelligence. The problem was a single structural omission: Johnson tested nothing before national rollout. He tried to rebuild JCPenney for a customer whose emotional relationship with the brand was inseparable from the couponing experience he was removing. One pilot store, for one quarter, would have revealed this before it became irreversible.

FAILURE 05  ·   JCPENNEY / RON JOHNSON   ·  The Transformation That Skipped Design, 2012–2013

The Design discipline requires testing the hypothesis before committing to a national scale — particularly when the strategy removes something the existing customer considers fundamental. Johnson’s reference group was wrong: he diagnosed based on Apple’s new-to-brand customers, not JCPenney’s decades-long coupon customers. Both levers were pulled simultaneously and nationally, with no test of their combined effect. It is the most instructive of the five failure cases because Johnson had the right instincts, the credibility, and the resources. What he lacked was the single discipline that would have caught the error before it became national, irreversible, and terminal.

OUTCOME:  Revenue fell 25% in year one, from $17.3B to $12.9B. Johnson terminated after 17 months. JCPenney filed Chapter 11 in May 2020.

These five cases share a common structure. The resources to act were present. What was absent was the discipline to price inaction honestly — to model what the competitive position would look like in five years if nothing changed. Both paths — refusing to act and acting without design — converge on the same outcome.

“The leaders in the success cases did not have more courage than those in the failure cases. They had more architecture. The discipline is the difference.”

THE DISCIPLINE IN PRACTICE

Five Leaders Who Structured Their Bets — and Won

The cases below are not simply success stories. They are demonstrations of the discipline in action — each one showing a different dimension of what it looks like to diagnose with rigor, design with structure, and decide with irrevocable commitment.

1.  Best Buy / Hubert Joly: Taking Amazon’s Bull by the Horns

When Joly arrived at Best Buy in September 2012, quarterly profits had fallen 91%. The consensus called for immediate store closures. Joly’s first act was diagnostic: he spent his opening days working on the floor of a Best Buy store — not in headquarters. What he observed was not structural failure but a pricing perception problem. Customers who had spent forty minutes with a knowledgeable associate were leaving because they assumed Amazon’s price was lower. In October 2012, Joly announced price-matching for any online competitor and named the strategy Renew Blue. “If you don’t have a name, you don’t have a plan, because people cannot relate to it.”

SUCCESS 01  ·   BEST BUY / HUBERT JOLY   ·  Renew Blue: The Price-Matching Gamble That Saved an Ecosystem

Calculated risk by the taxonomy: a written hypothesis (the conversion problem is perception, not price), a seasonal test before permanent rollout, a three-year recovery horizon modelled in advance, and a single named owner who staked his credibility publicly. The vendor partnership model — Samsung, Sony, Apple branded shop-in-shops — turned the showroom liability into a showroom advantage.

OUTCOME:  Best Buy shares rose from $11 at Joly’s appointment to over $65 at his exit — 263% growth. At least $1 billion in net profits every year from 2017. HBR named Joly one of the world’s best-performing CEOs in 2018.

2.  Williams-Sonoma / Laura Alber: Digital-First Before It Was Safe

In 2017, most home furnishings retailers were protecting their store networks. Laura Alber made the opposite call: digital-first, with stores recast as brand showrooms rather than primary transaction points. The bet rested on three structural advantages competitors had not acted on — proprietary in-house design that no marketplace could replicate, first-party customer data for personalisation, and a store estate that Amazon structurally could not build. Alber moved before being forced to. Williams-Sonoma was profitable and stable in 2017. The digital-first declaration was a choice, not a crisis response.

SUCCESS 02  ·   WILLIAMS-SONOMA / LAURA ALBER   ·  Digital-First in a Touch-and-Feel Category, 2017–2026

Calculated risk: the three structural advantages were diagnosed against a ten-year competitive horizon before capital was committed. The transition was staged — stores repositioned as showrooms, investment sequenced toward e-commerce, B2B developed in parallel. The strategy was named publicly and consistently, making it an organizational identity rather than a hedged experiment.

OUTCOME:  Approximately 70% of total revenue through e-commerce by 2026. EPS grew 14.4% year-on-year in fiscal 2024. Total shareholder returns over five years exceeded 180%.

3.  Abercrombie & Fitch / Fran Horowitz: Patience as the Most Disciplined Risk

When Horowitz became CEO in 2017, she inherited the brand the American Customer Satisfaction Index had named America’s most hated retailer. Revenue had declined for five consecutive years from a $4.5 billion peak. Her response was to choose patience as an explicit strategy. For five years she rebuilt the merchant function from the product level up, ran extensive consumer-listening programs, and invested in digital — all without announcing the destination. The public repositioning arrived in June 2022. Every element had already been rebuilt before the announcement was made.

SUCCESS 03  ·   ABERCROMBIE & FITCH / FRAN HOROWITZ   ·  From ‘Most Hated’ to Industry Benchmark, 2017–2023

Five years of designing before deciding, applied to one of the most complex brand repositioning challenges in recent retail history. Horowitz decided only when the execution infrastructure was fully in place, and announced only when the proof was complete. This is the direct reversal of the JCPenney error.

OUTCOME:  Record adult brand sales in 2023, growing at 6–8% CAGR. Named one of the most complete retail brand transformations of the decade.

4.  Pandora / Alexander Lacik: The Cleanest Illustration of the Design Discipline

When Lacik was appointed CEO in 2019, Pandora’s entire brand architecture rested on charm-collecting. His ambition — repositioning as a full jewellery company — was definitional in its risk. Lacik’s response was to expand on evidence rather than assertion. Lab-grown diamonds were introduced as a deliberate pilot test. Marvel and Disney collaborations tested new consumer cohorts. Each succeeded independently before the full repositioning was announced.

SUCCESS 04  ·   PANDORA / ALEXANDER LACIK   ·  Phoenix: Category Expansion via Pilot, 2019–2023

Calculated risk in its cleanest form: each category extension was a hypothesis tested at limited scale before scale investment. The full repositioning was announced only when the pilots had answered the brand extension thesis. The Phoenix strategy was named at the announcement and made irreversible, creating the commitment architecture that prevented post-decision drift.

OUTCOME:  Record revenue DKK 28.1 billion ($4.1B) in 2023 — 8% organic growth above guidance. Lab-grown diamond sales grew 116%. Pandora is the world’s most recognised jewellery brand.

5.  Walmart / Doug McMillon: People First, Then Platform

When McMillon became CEO in 2014, Walmart faced a structural problem no digital investment could solve first: a workforce averaging $8.81 per hour, with turnover so high that any omnichannel execution capability would be built on sand. In January 2015, he announced a minimum wage rise to $9 — then $10 in 2016 — at a cost of approximately $1 billion annually. Wall Street punished the decision immediately: shares fell nearly 10% in a single session, erasing roughly $20 billion in market capitalisation. McMillon’s hypothesis was that workforce stability was not a benefit of digital transformation. It was its prerequisite.

SUCCESS 05  ·   WALMART / DOUG McMILLON  ·  People Before Platform: The $1 Billion Bet Wall Street Punished, 2015–2024

Calculated risk by the taxonomy: a written hypothesis that workforce stability was the prerequisite for omnichannel execution, not its reward. A phased wage investment beginning at $9 per hour, rising to $10 in 2016, with the full digital bet sequenced behind it. A ten-year competitive horizon that priced the cost of inaction — specifically, the impossibility of building a same-day fulfilment network on a transient workforce — against the immediate margin compression Wall Street was pricing instead. McMillon absorbed the punishment of the short-term architecture willingly and publicly, because the Diagnose discipline had already answered the question the market was not asking: what does Walmart look like in 2025 if it does not solve this first?

OUTCOME:  Walmart U.S. e-commerce sales grew from approximately $12 billion in 2016 to over $75 billion by fiscal 2024. Walmart+ launched in 2020 and reached an estimated 32 million members by 2024. Employee turnover fell materially following the wage investment, enabling the store-as-fulfilment-hub model that now processes millions of curbside and same-day orders weekly. Walmart’s market capitalisation, which stood near $240 billion at the time of the wage announcement, exceeded $650 billion by 2024. The $20 billion erased on the day Wall Street punished the decision had compounded into the most significant value creation event in the company’s post-Walton history.

THE FRAMEWORK: Diagnose. Design. Decide.

How Ten Cases — and Thirty Years — Built a Framework for Structured Boldness

Frameworks rarely emerge from theory. This one did not. The Diagnose–Design–Decide structure was derived from the cases, and from thirty years of watching senior retail leaders across the Middle East, India, China, and emerging markets make consequential decisions under genuine competitive pressure.

One question was asked of every case: at the moment the critical decision was made, precisely which capability was absent that, if present, would have changed the outcome? Three answers emerged with enough consistency to constitute a pattern. In Circuit City, Blockbuster, and Sears, the failure was in how the decision was set up; inaction was never priced against the full strategic horizon. Walmart’s success case is the direct mirror: McMillon’s entire wage-first bet was built on exactly this discipline, modeling what Walmart’s digital position would look like in 2025 without a stable workforce before committing a dollar. In JCPenney, the failure was structural — the decision was made at full national scale without a test, when a single pilot would have revealed the fatal error. In Victoria’s Secret, the failure was in commitment — the diagnosis was complete, the decision was reached, but it was never named, never owned, never made irrevocable.

The success cases confirmed the pattern from the opposite direction. Every successful case applied all three disciplines. Every failure case had at least one structurally absent, built out of the organization’s decision-making process at the moment it was most needed.

DIAGNOSE	DESIGN	DECIDE
Price inaction across the full strategic horizon — not just the current cycle. Map both action risk and inaction risk with equal rigour. Pressure-test assumptions with external perspectives. In practice: Joly spent his first days on a Best Buy shop floor watching showrooming happen before he prescribed the cure.	Pilot before committing. Write the hypothesis first. Define explicit success metrics and the threshold that triggers scale. Never pull two levers simultaneously without testing their combined effect. In practice: Lacik tested lab-grown diamonds in one market before betting the Pandora brand on fine jewellery.	Name the bet publicly — a strategy without a name cannot be held. Commit irrevocably when the pilot answers the hypothesis. Build a formal learning loop from Day 1. In practice: Horowitz rebuilt Abercrombie silently for five years, then announced only when every element of execution was already in place.

The Framework From the Inside

Wansa at Xcite Kuwait: When the Author Lived Every Step

The ten cases above were observed from the outside — after the outcomes were known, after the metrics had moved, after history had rendered its verdict. What follows is different. It is the same discipline applied from within: before the outcome was known, before a single internal or external voice believed the bet could work, and under conditions that made abandonment look rational on every spreadsheet for three consecutive years. I am placing it here, after the framework and separate from the external cases, deliberately. A leader who applies this framework to their own organisation deserves to know that the author has done the same.

The consumer electronics market in Kuwait in the early 2000s was, by any analysis, the wrong place to build a private label. Samsung, LG, Sony, and Panasonic dominated every category. The Kuwaiti consumer was brand-literate and brand-loyal — and had every reason to be. The consensus view, inside the organization and among our commercial partners, was unanimous: no private label could compete in this environment. That consensus was wrong.

Diagnose: Pricing Inaction Honestly

In 1999, Alghanim Industries’ Chairman asked his team to explore whether a private label could improve margins. I was tasked with making it happen from China — one person, no language, no contacts, no internet, landing in Beijing with a vague mandate. We were told within hours that we had come to the wrong city. Electronics manufacturing was in Qingdao and Xiamen. A week was lost meeting suppliers who held no export licenses.

The Diagnose question we asked was not “can a private label work here?” It was more specific: what would be true about Xcite’s margin structure and pricing power in ten years if we remained entirely dependent on Samsung, LG, and Sony? That question made inaction expensive. It made the risk of trying far smaller than the risk of not trying.

Design: The Pilot That Survived Failure

At the Canton Fair we placed our first orders: $100,000 in consumer electronics with written voltage and quality specifications. Three months later, every unit failed. Products built for China’s domestic grid did not conform to 240V. Customers returned them. It was a complete write-off. Most organisations would have ended the experiment there. Our owners told us to find a better way.

The Design response was specific. Written specifications were not enough — I relocated to China. In June 2002 we opened a sourcing office in Guangzhou: one person growing to a team of thirty local staff, relationships with over 100 suppliers across 40 cities, more than 1,000 factory visits. We built a quality manual for every product category and stationed inspectors on factory floors before and after assembly. We scoped the pilot to small domestic appliances — lowest brand loyalty, highest measurability — and named three criteria for progression: return rate, repurchase rate, and net promoter score.

The brand damage ran deeper than the technical fix. Wansa’s reputation had collapsed so completely that customers refused the products even as free gifts. The quarterly case for abandoning the brand looked rational on every spreadsheet. We held — not through stubbornness, but because the hypothesis metric we had committed to track — the repurchase rate among customers who had experienced no failure — was moving in the right direction. When small houseware quality refused to reach threshold after sustained effort, we made the hardest call the discipline requires: we exited that category entirely rather than allow further failures to poison the recovery in others.

Decide: Naming the Commitment Irrevocably

By year three, we expanded into air conditioners and larger appliances — the categories where the brand premium was highest and our value proposition most powerful. The Decide discipline required naming the commitment publicly and making it irrevocable: dedicated floor space, Wansa-trained sales staff, and a marketing programme positioned on value-for-quality rather than discounting. We staked the organisation’s credibility on the bet before the income statement justified it.

By 2006, Wansa was the market leader in Kuwait by sales volume, outselling Samsung, LG, Sony, and Panasonic in its core categories. By 2009, the China sourcing operation had become a $500 million platform — 25% of Alghanim Industries’ revenue and 50% of group profit.

WANSA / XCITE BY ALGHANIM ELECTRONICS   ·  Building Consumer Electronics’ Most Unlikely Private Label, 2001–2010

Diagnose: The hypothesis identified a specific, underserved segment — value-seeking mid-tier appliance buyers — and mapped the competitive landscape against a ten-year horizon. The question asked was not whether a private label could succeed, but what Xcite’s position would look like in a decade without one. Design: A dedicated sourcing office in Asia was established before any significant product launch. Three scaling criteria were named up front: return rate, repurchase rate, net promoter score. Two years of below-target financials did not trigger abandonment, because the hypothesis metrics were moving correctly. One underperforming category was exited entirely rather than allowed to contaminate the recovery. Decide: The commitment was named and made irrevocable through dedicated floor space, trained staff, and a public marketing commitment to the Wansa brand — before the income statement justified it.

OUTCOME:  Wansa became the largest-selling individual brand at Xcite Kuwait by sales volume — outselling Samsung, LG, and Sony in its core categories. Today it carries over 1,800 active products, exceeds $300 million in brand revenue, and has expanded into Saudi Arabia. No consumer electronics private label in the Gulf had achieved this position before. The consensus that it could not be done was not wrong about the market. It was wrong about what the discipline makes possible.

THE PSYCHOLOGY OF CAUTION

“When inaction is riskier than the action”

Why Capable Leaders Apply the Wrong Discipline

If the Diagnose–Design–Decide structure is this clear, why do board-approved leaders consistently fail to apply it at the most critical moment? In thirty years of working with senior retail leaders, I have observed three consistent dynamics that cause even rigorous organizations to substitute the wrong discipline for the right one.

1. The Asymmetry of Visible and Invisible Risk

The risk of bold action is always visible — it can be modelled, scenario-planned, and presented with a probability distribution. The risk of inaction is invisible. It accumulates slowly, off the income statement, in competitive ground ceded and customer relationships quietly migrating. Sears’ board did not choose bankruptcy in 2003. They chose not to invest in digital capability in 2003. Behavioural economics calls this omission bias — the systematic tendency to judge harmful actions as worse than equally harmful inactions. In retail boardrooms, it manifests as the consistent under-pricing of standing still.

2. The Short-Term Architecture of Retail Governance

Most retail boards evaluate management against quarterly comparable sales, gross margin, and EBITDA. None of those metrics capture the cost of a digital capability not built, a consumer cohort not reached, or a brand identity not refreshed. The incentive architecture rewards the discipline of optimisation — which is the right discipline for a stable competitive environment, and precisely the wrong one for a disrupted category. The CEO who protects margin by deferring transformation is rewarded in the same quarter the damage is done. The CEO who invests in a bet with a three-year payback horizon absorbs the cost immediately and the reward later — often under a different CEO.

3. The Competence Trap

The leaders who built the businesses being disrupted were good at the model being disrupted. Ron Johnson built the Apple Store. Sears’ executives were among the most experienced operators in American retail. Their expertise was real — and it was calibrated to a world that was changing. Horowitz understood this instinctively. She spent five years deliberately developing the judgement of a consumer she had never previously served, before committing to a single public decision. The patience was not a suspension of discipline. It was the construction of the right one.

THE MANDATE AHEAD

 “"The biggest risk is not taking any risk... In a world that's changing really quickly, the only strategy that is guaranteed to fail is not taking risks"- Mark Zuckerberg, Founder, Meta

The Discipline of Calculated Risk

In the decade ahead, the discipline of calculated risk will become the primary competitive differentiator in retail. AI is compressing decision cycles from weeks to hours. Shein and Temu demonstrated that scale is no longer a structural moat — they built global supply chains faster than legacy retailers could schedule the board meeting to discuss the threat. Gen Z, who by 2030 will represent the largest consumer segment in developed markets, evaluates brands not on stated values but on demonstrated decisions. A retailer that has not made a publicly named, committed, structurally disciplined bet on its own future by 2027 will not read as cautious to this cohort. It will read as irrelevant.

There is a misreading of this that deserves direct confrontation. AI is accelerating the Design step — pilots that once took eighteen months can now be instrumented and read in sixty days. Some leaders will conclude that this makes the Decide discipline less critical: if testing is faster, commitment can wait longer. The logic inverts. When every competitor has access to the same AI-driven testing infrastructure, speed of commitment becomes the only durable differentiator. The retailer that runs the fastest pilots but hedges the commitment has paid for the intelligence and refused the advantage. Lacik did not pilot lab-grown diamonds because the technology was slow. He piloted because the hypothesis had to be proved before the bet could be named irrevocably. The speed of the pilot was never the point. The irreversibility of the commitment was. AI makes the Design step faster. It makes the Decide discipline more consequential, not less.

The Diagnose–Design–Decide framework is not a guarantee of success. What it eliminates is the far more common and far more costly failure mode: arriving at the end of the strategic window having never applied the rigour that would have told you which bet to make, when to make it, and how to make it survivable.

The leaders in the failure cases were not undisciplined people. They were disciplined people whose organizations had built the wrong discipline — optimised for certainty, rewarding the avoidance of visible risk, and systematically blind to the compounding cost of the risks they refused. Building the right discipline is the real leadership work. Not the strategy review. Not the consultancy engagement. Not the board presentation. Those are outputs of the discipline. The discipline itself is what this article has been about from the first line.

“The leaders who failed were not undisciplined. They had built the wrong discipline — one that priced the risk of acting, and left the compounding cost of inaction entirely off the ledger.”

THREE QUESTIONS TO TEST YOUR DISCIPLINE

1.  For your three most consequential decisions in the past two years: what was the formal analysis of the cost of not acting? If none exists, that is the first capability to build.

2.  What transformation has your organization been diagnosing without piloting for more than eighteen months? Name the pilot you could run in ninety days for under 1% of the full commitment budget.

3.  What is the name of your organization’s single most important strategic bet for the next three years? If your executive team cannot say it consistently and without hesitation, the commitment does not yet exist.

Trust us to get your leaders to be at their best!

Connect with us

Know your mentor