Gather round. It is late October, the nights are drawing in, and there is something deeply appropriate about sitting down with a strong coffee and a set of business decisions that did not end well.
These are not fictional horror stories. No monsters, no fog machines. The truly frightening thing about what follows is that every one of these situations was, in some sense, avoidable. Not easily. Not obviously, in the moment. But avoidable nonetheless. Each one carries a lesson that is still relevant today, for any founder, managing director or executive sitting in front of a difficult decision and wondering whether they are asking the right questions.
Consider this a seasonal reminder that the most dangerous threats to a business rarely announce themselves dramatically. They accumulate quietly, while everyone is looking somewhere else.
Blockbuster: the danger of mistaking loyalty for permanence
In the early 2000s, Blockbuster was not a struggling business. It had thousands of stores, millions of members and a brand so embedded in weekend culture that it felt institutional. When Netflix arrived, offering DVD rentals by post without the late fees that Blockbuster customers quietly resented, the response from Blockbuster's leadership was, broadly speaking, indifference.
The logic was understandable. Why cannibalise a profitable estate of physical stores for an unproven mail-order model? Why respond to a competitor with no track record in the industry?
Because the customer was already answering that question on behalf of the board.
By the time streaming became viable and Netflix pivoted again, the window for Blockbuster to respond had essentially closed. It filed for bankruptcy in 2010. There is now one Blockbuster store left in the world. It is in Oregon, and it sells merchandise.
The lesson is not that Blockbuster failed to predict the future. It is that the signals were present in the customer behaviour, in the complaints about late fees, in the early adoption numbers for Netflix, and those signals were not treated as strategically important. Listening to customers is not a soft skill. It is an early warning system.
Kodak: inventing the thing that ended you
Here is a detail that still lands oddly when you first hear it. Kodak's own engineers developed a working digital camera in 1975. The company held patents that put it at the centre of digital photography. It understood, technically, what was coming.
It chose not to pursue it aggressively because digital threatened the film and printing revenue that made Kodak profitable. This is sometimes called the innovator's dilemma, and it is a real and well-documented phenomenon: the more successful a business is in its current model, the harder it is to invest properly in the thing that will eventually replace it.
Kodak filed for bankruptcy in 2012.
What makes this particular story instructive is the distinction between knowing something and acting on it. Kodak had knowledge. What it lacked was the organisational willingness to treat that knowledge as a strategic priority over the objections of the people whose divisions it threatened. That is a governance problem, not a technology problem. The camera was not the issue. The room where decisions got made was the issue.
New Coke: when you solve the wrong problem
In 1985, Coca-Cola reformulated its flagship product after internal research showed that consumers, in blind taste tests, preferred a sweeter alternative. The logic was clean. The data supported the decision. New Coke launched.
The backlash was immediate and furious.
Coca-Cola's research had, rather fatally, failed to account for the emotional relationship customers had with the original product. The taste test measured preference in isolation. It did not measure what it meant to customers that the original Coke was being discontinued. Those are different questions, and only one of them turned out to matter.
Market validation is not the same as customer insight. You can construct a research method that answers the question you asked perfectly and still miss the question that actually needed asking. The company reversed course within three months, relaunching Classic Coke to significant relief and, it has to be said, rather extraordinary sales uplift that may or may not have been entirely unplanned.
The lesson for anyone working on a product change, a pricing shift or a new go-to-market strategy is this: be deliberate about what your research is actually measuring, and curious about what it is not.
Nokia: being good at the old thing
Nokia made extraordinarily well-built phones. Its hardware quality was genuinely respected, its market share was dominant, and its internal engineering culture was strong. What it underestimated, and there is now substantial academic and journalistic record on this, was the degree to which the smartphone would redefine what a phone actually was.
The iPhone was not a better Nokia. It was a different category of device that happened to make calls. Nokia's engineers understood hardware. The shift required a different kind of thinking: software ecosystems, developer relationships, platform strategy.
There is also a documented internal culture problem. Research published after Nokia's decline described an organisation in which junior employees were reluctant to tell senior leadership about problems. Managers feared delivering bad news upward. The result was that accurate information about competitive threats did not reach the people who needed it in a form that prompted action.
Nokia's handset division was acquired by Microsoft in 2014 for a fraction of what the business had been worth at its peak.
A culture that discourages dissent does not just feel uncomfortable. It makes the organisation functionally blind to the things it most needs to see. That is worth remembering every time someone in a leadership team says they operate an open-door policy.
Target Canada: moving faster than your foundations
Target's expansion into Canada in 2013 is one of the more instructive operational disasters of recent memory, and it is instructive precisely because the failure had almost nothing to do with brand awareness or consumer appetite. Canadians knew Target. Many had shopped at American stores across the border and were genuinely enthusiastic about the arrival.
What collapsed was the supply chain.
Target Canada's inventory systems were, reportedly, full of inaccurate data from the start. Product dimensions, weights and supplier codes were riddled with errors inherited from the acquired store estate. The consequence was that shelves sat empty or were stocked inconsistently, customers came in expecting a full retail experience and found gaps, and the brand reputation began to erode almost immediately.
The business had grown faster than its operational foundations could support. Target Canada filed for creditor protection in January 2015, less than two years after opening its first stores. One hundred and thirty-three stores were closed.
Growth ambition without operational readiness is not bold. It is expensive. For any SME considering a new market, whether geographically or by sector, the question of whether the back-end infrastructure can actually support the front-end promise is not secondary. It is the thing to stress-test before you commit.
Equifax: the cost of leaving the door open
In 2017, Equifax, one of the world's largest credit reporting agencies, disclosed that attackers had accessed the personal data of approximately 147 million people. Names, dates of birth, addresses, Social Security numbers, and in some cases driving licence and credit card details.
The breach had exploited a known vulnerability in a web application framework, one for which a patch had been available for months. The patch had not been applied.
The consequences were significant, including fines, legal settlements and reputational damage. A company whose entire value proposition rested on the secure handling of sensitive personal data had, essentially, left a well-documented door unlocked.
Cybersecurity failures at this scale are sometimes treated as sophisticated, inevitable or technically incomprehensible by those who experience them. Many are not. A significant proportion of breaches, including some of the most damaging, begin with something comparatively basic: an unpatched system, a weak password, a phishing email that reached the wrong inbox, a supplier with access that was never reviewed.
The unglamorous truth is that basic hygiene prevents a substantial share of attacks. The Equifax case is a useful reminder that cybersecurity risk is not only about the sophistication of adversaries. It is also about the discipline of maintenance. For any business handling customer data, that question deserves a direct answer, not a reassuring assumption.
Theranos: when governance fails, fraud fills the gap
Elizabeth Holmes founded Theranos on a genuinely compelling idea: that a small drop of blood could be used to run hundreds of diagnostic tests cheaply and quickly, making healthcare more accessible. The pitch attracted serious investors, board members with significant reputations, and coverage in major publications.
The technology did not work as described. The company used conventional machines for most of its tests while presenting itself publicly as having solved a different kind of problem entirely. Holmes was convicted of fraud in 2022.
There are several things to take from Theranos, and not all of them are obvious. The board was distinguished but lacked anyone with relevant technical expertise who might have asked the right questions. The culture, by multiple accounts, actively discouraged internal challenge. And the investors, including those with substantial experience, failed to conduct the kind of rigorous due diligence that the scale of their commitment warranted.
Theranos is often discussed as a story about one person's deception. It is also a story about what happens when governance is decorative rather than functional, when the people responsible for asking hard questions do not ask them, or are not positioned to do so.
The practical question for any board, leadership team or investor is not whether they trust the founder. It is whether they have created conditions in which they would actually find out if something was wrong.
What to do with all of this
These stories span different industries, decades and failure modes. But they share a set of recurring conditions: leadership teams that were not hearing what they needed to hear, organisations that moved faster than their capabilities could support, assumptions that were tested too late or not at all, and governance structures that looked adequate until they weren't.
None of this requires you to become paranoid. It requires something more useful: curiosity about where your own blind spots might be.
A reasonable set of questions to sit with, perhaps while the Halloween decorations are still up:
- Where in your business are people unlikely to tell you something is going wrong?
- When did you last validate a core commercial assumption against actual customer behaviour, not internal consensus?
- If your main operational system failed tomorrow, how long before a customer noticed, and how long before it became a crisis?
- Who in your leadership team is responsible for asking the uncomfortable question, and do they actually ask it?
- Is your cyber housekeeping current, or is there a patch somewhere that nobody quite got round to applying?
Fear, in a business context, is most useful as foresight. The stories above were not inevitable. They were the product of decisions that could have been made differently, at moments when the information to make them differently was, in most cases, already available.
That is the thing about horror stories. The clues are usually there from the beginning. You just have to be paying attention.
