Reads Revisit list

← Back to Issue

Zombie companies

From Leadership in Tech · subscribed via aiste.ulozaite@gmail.com · original ↗ · unsubscribe

Large companies can keep growing, renewing contracts, and reporting strong results long after their ability to build a relevant future has gone. Past advantages like brand recognition, contracts, and switching costs keep money flowing, but they say nothing about whether the business could earn its position again today. Meanwhile, challengers need less capital than ever to pick apart the most valuable pieces, and AI is making it cheaper for customers to inspect, compare, and leave.

Zombie companies

14th July, 2026

Some companies fail years before their results reveal it.

They continue to grow, renew contracts, acquire competitors and announce record quarters. Their share price holds. Their market share looks reassuring. Internally, the dashboards are green, the board is pleased, and the strategy appears to be working.

But something more fundamental has already shifted – the organisation is no longer shaping the conditions that will determine its future.

That failure rarely looks dramatic when it happens. There is no single catastrophic decision, no obvious moment where the company crosses from healthy to doomed. The product still sells. Customers still tolerate it. The brand still carries weight. Sales teams still close deals on the strength of reputation, relationships and procurement inertia. The machinery keeps moving, so everybody assumes the engine is still running.

This is how companies become zombies.

Table of contents

They remain commercially alive because the market is still rewarding decisions made years or decades earlier. Brand recognition, distribution, contracts, accumulated authority, switching costs and sheer scale continue to push money through the system. Those advantages are real, but they say very little about whether the organisation could earn the same position again under today’s conditions. Current success is often historical success arriving late.

That distinction is easy to miss because businesses measure what flows through them, rather than what made those flows possible. Revenue tells you that customers paid. Retention tells you that they stayed. Market share tells you that competitors have not yet displaced you. None of those metrics reveal whether customers would choose you from scratch, whether your category is becoming less important, or whether your organisation can build whatever comes next.

A company can therefore look successful, feel successful and be rewarded as successful, while its strategic trajectory has already collapsed. The failure exists before the evidence becomes financially convenient enough to acknowledge.

The numbers are not lying; they are simply looking backwards. And the larger the company, the longer it can continue. Accumulated advantage creates momentum, and momentum can sustain performance long after capability has decayed.

Success is a lagging indicator

Most companies are judged by what they have already managed to accumulate.

Revenue, margin, market share, customer count, brand recognition, distribution, authority, institutional knowledge, political access, contractual lock-in, procurement familiarity and years of embedded habit are all treated as evidence of strength, whether or not accounting recognises them as assets. For a long time, they were.

And to be clear, they still are. A famous brand opens doors. A large installed base creates predictable income. Deep integrations make customers reluctant to leave. Capital buys time, attention and optionality. Regulatory expertise keeps smaller competitors locked out of the room. Distribution ensures that the incumbent is seen first, stocked first, recommended first and often chosen by default.

The problem begins when those advantages are mistaken for proof of continuing capability. Most of them describe what the company has built up, rather than what it can still build. They tell you how much force remains in the system, but very little about whether the organisation is generating any new force of its own.

That is the danger of lagging indicators. They are excellent at describing the performance of the system which already exists, and almost useless at telling you whether that system remains capable of creating its successor.

A healthy organisation uses the proceeds of past success to build new sources of advantage. A failing one uses them to preserve the appearance that nothing fundamental has changed.

From the outside, both can look successful for years. The difference is hidden in what the organisation is still capable of becoming.

What makes a company a zombie?

The defining feature is a growing separation between what the company owns and what it can still do.

It may own a famous brand, a large customer base, valuable data, entrenched distribution, expensive infrastructure, deep regulatory knowledge and decades of accumulated trust. Those assets matter. They create access, stability and protection from mistakes which would kill a smaller business.

But assets are not the same thing as capability.

Capability is the organisation’s ability to recognise when the market is changing, make consequential decisions, build something which threatens its existing model, and move resources towards an uncertain future before the current one becomes visibly untenable. It is the capacity to turn advantage into renewal.

A company is not a zombie merely because it is old, bureaucratic, shrinking or harvesting a mature product. Those can all be rational conditions or choices. The zombie state requires three things to be true at once: current performance depends heavily on advantages accumulated under earlier conditions; the sources of future demand are moving elsewhere; and the organisation cannot follow them without damaging the business it is rewarded for protecting.

Without all three, this may simply be maturity, specialisation or ordinary decline. With them, the company can remain healthy on paper while its viable futures quietly disappear.

The company can still operate the machinery it inherited. It can optimise pricing, improve margins, launch incremental features, expand into adjacent markets and extract more value from established relationships. It may be exceptionally good at running the business which already exists.

What it cannot do is create the business which will need to exist next.

That distinction is easy to miss because operating competence looks like strategic health. A company which executes reliably, forecasts accurately and delivers predictable returns appears well governed. Its systems reduce variance, control risk and make the future resemble the past.

Those are sensible qualities until the future stops cooperating.

At that point, the organisation discovers that everything which made it reliable also made it rigid. New ideas must fit existing departments. Emerging products must justify themselves using the economics of mature ones. Alternatives look naive, incomplete or commercially irresponsible because they threaten the assumptions through which the company understands success.

The organisation still has resources, but those resources have become difficult to use for anything which challenges its own logic.

A resilient incumbent uses brand, capital, distribution and expertise to widen the range of futures available to it.

A zombie uses those advantages defensively. It acquires threats rather than learning from them. It uses distribution to keep weaker products visible, contracts to prevent movement, brand recognition to postpone improvement, and scale to absorb inefficiency.

Eventually, it becomes incapable of imagining success which does not preserve its current products, margins, hierarchy and importance. Every proposed future must protect the present, even when the present is precisely what needs to change.

By then, the organisation may already be dead in the only sense that matters. It can continue doing what it has always done, but it can no longer become something else.

Delay is mistaken for resilience

Markets rarely punish strategic failure on schedule.

Customers do not all wake up on the same morning and decide that a product has become obsolete. Contracts expire gradually. Procurement teams move cautiously. New competitors take time to earn trust. Habits survive long after better alternatives appear. Entire categories can continue generating enormous amounts of money while the reasons for their existence quietly weaken underneath them. That delay is dangerously reassuring.

A company sees a new competitor gaining attention, watches it fail to make an immediate dent in revenue, and concludes that the threat was overhyped. Customers complain but continue renewing, so the complaints are treated as noise. New behaviours emerge among smaller or younger audiences, but the existing customer base remains large enough to make them look peripheral. The market shifts at the edges while the centre appears stable.

Leadership looks at the numbers and sees resilience. What it may actually be seeing is latency.

Large organisations are protected by countless sources of delay. A familiar brand reduces the perceived risk of choice. Long contracts postpone customer decisions. Deep integrations make migration painful. Internal champions defend established suppliers because changing course would expose their own earlier judgement. Entire teams, workflows and reporting structures may be built around the incumbent’s product, turning dissatisfaction into a problem which feels more expensive to solve than to tolerate. 

All of this buys time. Time, which is useful when an organisation spends it adapting, but which becomes lethal when interpreted as evidence that adaptation was unnecessary.

The distinction matters because a moat and momentum can look almost identical from inside the company. Both produce continued sales, persistent market share and customers who remain in place. Both make competitors appear weaker than the incumbent. Both reassure investors that the business remains difficult to displace.

But a moat makes the company harder to attack. Momentum merely makes the consequences of attack slower to appear.

A genuine moat continues to protect because the underlying advantage remains active. Better data can improve the product. A larger network can attract more participants. Scale can lower costs. Trust can deepen through continued delivery. Success continues to reinforce the mechanism which produced it.

Momentum behaves differently. The company continues moving because it already has weight, velocity and a long stretch of road behind it. Brand recognition still opens doors, but less reliably. Customers still renew, but with more resentment. Distribution still provides reach, but emerging channels are forming elsewhere. Switching costs still hold people in place, but migration is becoming easier.

The outward result remains positive while the underlying forces weaken.

That gap between cause and consequence creates a profound political problem inside the organisation. People who warn about the trajectory are arguing against visible evidence. They are asking leaders to damage profitable products, redirect resources and accept near-term uncertainty in response to threats which have not yet shown up in the financial results.

The defenders of the current strategy have revenue, renewals and market share on their side. The challengers have patterns, weak signals and an increasingly uncomfortable theory about the future. Corporate decision-making tends to reward the people holding the spreadsheet.

So the company waits for proof. It waits for churn to rise, margins to fall, competitors to scale, and for customers to demand change loudly enough that nobody can dismiss them. It insists on evidence which will only become conclusive after the window for an easy response has closed.

Then, when the numbers finally turn, the organisation behaves as though the problem has just arrived. It launches reviews. It commissions research. It replaces leaders. It announces a transformation programme. Everybody suddenly agrees that the business must become faster, simpler, more innovative and more customer-focused.

The urgency is real, but (too) late. The market has spent years learning how to live without the company, while the company was congratulating itself for surviving another quarter.

A zombie organisation can remain in this state for a surprisingly long time. Its accumulated advantages soften every blow, stretch every decline and postpone every reckoning. Each delay is interpreted as another victory, even as the trajectory becomes harder to reverse.

Eventually, that habit of misreading the evidence begins to shape the organisation itself. The company learns which signals to ignore, which questions to discourage, and which kinds of activity create the appearance of progress without threatening the existing model.

That’s when the symptoms become difficult to miss.

The symptoms of organisational undeath

Zombie companies rarely look neglected. They are often polished, active and well-managed.

They launch products, publish strategies, announce partnerships, hire senior leaders, reorganise departments and produce endless evidence that important work is happening. Their offices are full of intelligent people solving difficult problems. Their calendars are packed. Their roadmaps are ambitious. Their internal communications are relentlessly optimistic.

The problem is that most of this activity is directed towards preserving the appearance of health, rather than restoring the capacity to change.

They protect the evidence of success

As growth becomes harder to create, the organisation becomes increasingly skilled at protecting the numbers which say that growth is still happening.

Prices rise. Products are bundled. Previously standard features are moved into premium tiers. Contracts become longer. Cancellation becomes harder. Support becomes cheaper to provide and more frustrating to use. Sales teams focus on expanding existing accounts because acquiring genuinely new customers has become expensive and uncertain.

None of this looks like decline in a quarterly report. Quite the opposite. Revenue per customer improves. Margins expand. Retention remains reassuringly high. Leadership points to the results as evidence that the strategy is working.

But the business is no longer creating proportionate amounts of new value. It is improving its ability to extract value from a position created earlier.

That distinction becomes particularly easy to ignore when the customer base is large and embedded. A small increase in pricing across millions of accounts can produce more immediate growth than years of uncertain product development. Removing costs from support can improve margins faster than improving the experience. Making a bundle harder to escape can outperform building products customers would actively choose to keep.

Each decision is defensible in isolation. Collectively, they reveal an organisation consuming trust, goodwill and flexibility to preserve financial performance. The company becomes more profitable and less useful at the same time.

Zombie companies become experts in measuring the harvest while avoiding questions about the soil.

They perform change without changing

The most visible symptom is the reorganisation.

A new chief executive arrives. Business units are consolidated. Product teams are reshaped around customer journeys. Regional structures become global, then global structures become regional again. Centres of excellence appear. Transformation offices are funded. Decision-making is supposedly pushed closer to the customer, while three new layers of oversight are created to make sure that happens consistently.

The diagrams change. The organisation does not.

Reorganisations are attractive because they turn strategic anxiety into administrative activity. They create new titles, new reporting lines, new meetings, new objectives and an immediate sense that leadership is taking control. They can be announced, tracked and completed. They provide a visible response to a problem whose real causes are politically dangerous and operationally messy. 

Changing what the company is for is much harder. That might require abandoning profitable products, removing executive territory, rewriting incentives, closing divisions, accepting cannibalisation or admitting that years of investment were directed towards the wrong future. It might require fewer committees, fewer management layers and fewer people whose careers depend on the complexity they oversee.

An org chart offers a safer target. So teams are moved without gaining authority. Responsibilities are redistributed without removing vetoes. New leaders inherit the same targets, dependencies and constraints as the leaders they replaced. The company creates cross-functional groups to navigate problems created by its functional silos, then creates programme-management layers to coordinate the cross-functional groups.

Eventually, an extraordinary proportion of the organisation exists to manage the difficulty of operating the organisation.

And so progress becomes increasingly theatrical. Transformation is narrated through town halls, strategy decks and internal slogans. Programmes are named, branded and launched with great ceremony. Every initiative promises simplicity, speed and customer focus, while introducing more governance to ensure that simplicity, speed and customer focus are delivered correctly.

When a company can no longer change its trajectory, it changes its org chart. But the underlying commercial assumptions remain untouched. The company still needs to protect the same revenue, satisfy the same markets, support the same legacy, preserve the same hierarchy and avoid the same uncomfortable trade-offs. It changes shape without changing direction.

They defend the old market

Zombie organisations tend to be highly sophisticated when evaluating competitors and curiously unimaginative when interpreting them.

A new entrant is examined against the standards of the mature incumbent. It lacks features. Its product is shallow. Its support model would never satisfy enterprise customers. Its security is unproven. Its margins look unsustainable. Its users are hobbyists, small businesses, teenagers, creators or some other group which respectable market analysis can safely place at the edge.

Every criticism may be correct. But what the incumbent misses is that the challenger does not need to reproduce the current market in miniature. It may be building for a different version of the market entirely.

The company compares products as they exist now, rather than comparing the directions in which they are moving. It asks whether the newcomer can serve its largest customers today, rather than why an increasing number of people prefer the newcomer’s assumptions about simplicity, pricing, access or control.

That mistake becomes particularly dangerous when the challenger begins by serving customers the incumbent does not especially want.

Low-value users are encouraged towards cheaper tools. Small businesses are dismissed as uneconomical. New workflows are treated as unsophisticated. Simpler use cases are surrendered because they do not justify the complexity or pricing of the incumbent’s product.

This might look sensible – the company is focusing on its most profitable customers. But entry-level products, informal workflows and marginal users often determine where habits form. They shape what new employees expect, what new businesses adopt, what educators teach, what communities recommend and what future buyers consider normal. 

The incumbent retains the most valuable customers while gradually losing control of how customers become valuable. And its response is usually to absorb the surface features of the threat. It launches a lighter tier, adds an AI assistant, simplifies a dashboard, creates templates, announces a creator programme or acquires a fashionable start-up.

But those interventions are required to reinforce the existing model. The cheaper tier must lead customers towards the premium one. The AI assistant must make the established workflow more efficient. The acquired product must integrate with the portfolio. The new audience must eventually behave like the old audience. The company keeps improving the answer after the market has begun asking a different question.

They lose the future while retaining the present

The most important losses usually begin among customers who contribute almost nothing to current revenue.

A student learns a different tool. A small company adopts a simpler platform. A junior employee brings a new workflow into the organisation. A creator builds an audience somewhere the incumbent barely measures. A new business starts without the systems, habits or assumptions which made the established product feel unavoidable.

Individually, these people are commercially insignificant. Collectively, they determine what the market will expect next.

Zombie companies struggle to see this because their understanding of importance is shaped by the current revenue base. Large enterprise customers matter. Mature markets matter. Established purchasing processes matter. The people who are already paying matter. New behaviour tends to look small precisely because it is new.

The company therefore retains its largest accounts while becoming less relevant to the people entering the category. Its customer base grows older, more embedded and more expensive to support. Its brand remains famous, but the fame changes character. It becomes something people recognise rather than something they aspire to use.

The organisation may respond with increasingly aggressive declarations of innovation. Its corporate language grows grander as its relationship with reality weakens. Every release becomes transformational. Every partnership reshapes the industry. Every modest improvement demonstrates leadership. Criticism is attributed to poor communication, customer misunderstanding or resistance to change.

The company becomes exceptionally good at narrating progress which its customers no longer experience.

Inside the organisation, that narrative creates its own politics. People who present good news gain access and influence. People who identify structural problems are told to be more constructive, more commercial or more solutions-oriented. Weak signals are softened as they travel upwards. Problems become risks, risks become challenges, and challenges become opportunities.

Eventually, senior leadership is surrounded by extensive evidence that the business is changing, produced by an organisation which has learned that describing change is safer than causing it.

The organisation does not remain still. It moves constantly, with enormous energy, around the boundaries of the future it is unwilling to enter.

By this point, the obvious question is why somebody does not simply intervene. These companies have capital, talent, data, distribution and every conceivable strategic warning. They can see the challengers. They can hire consultants. They can acquire technology. They can read the same market signals as everybody else.

The uncomfortable answer is that recognising the future and being able to move towards it are entirely different capabilities. By the time an organisation exhibits these symptoms, the current business may have made meaningful adaptation structurally impossible.

Why the company cannot simply adapt

By the time a company recognises that its market is changing, the advice is usually some variation of “adapt faster”.

Invest in innovation. Empower smaller teams. Modernise the platform. Launch a new product. Buy a start-up. Add AI. Become more customer-focused. Move with urgency.

A large incumbent should be capable of all of those things. It has money, talent, reach, data and access to expertise. It can see the threat, describe the future and commission several hundred slides explaining what must happen next.

What it often lacks is permission to follow the argument to its conclusion.

Meaningful adaptation usually requires choosing against something which already works. The new product may need to be cheaper than the old one. The new workflow may remove services which currently generate revenue. The new market may be smaller or less profitable. The new business may require different skills, fewer people, unfamiliar distribution or the abandonment of customers whose needs shaped the organisation for years.

Somebody loses.

A division loses revenue. An executive loses authority. A team loses headcount. A product loses priority. A partner loses influence. Investors lose short-term certainty. Customers lose features or workflows which the company spent years promising to support.

That is where adaptation becomes political.

The organisation can agree enthusiastically that change is necessary while every part of it behaves rationally to prevent change in practice. Finance protects margin. Sales prioritises deals it can close. Product avoids destabilising valuable customers. Technology worries about legacy dependencies. Legal expands the cost of experimentation. The board demands accountability for capital.

Each function protects the company from immediate harm. Together, they protect it from becoming anything else.

Nobody needs to reject the future outright. They only need to ask sensible questions until it arrives too late.

The existing business has already defined what good judgement looks like. Investment cases are built around current margins. Opportunities are assessed through current customer segments. Performance reviews reward current revenue. Risk models assume continuity. Strategic proposals gain credibility when they resemble previous successes.

The future must therefore justify itself using the logic of the past.

A fragile new product will always look commercially weak beside one with years of optimisation behind it. A simpler proposition looks underpowered beside an enterprise suite. A small emerging segment looks irrelevant beside a large installed base. A disruptive pricing model looks irresponsible beside predictable recurring revenue.

The organisation compares a beginning with an end state, then congratulates itself for choosing the safer option.

This is why structural threats produce peripheral experiments. A small innovation team is allowed to explore, provided that it does not disturb the core business. A new product is launched under a separate brand, but must use the existing systems, targets and procurement processes. A start-up is acquired for its speed and originality, then subjected to governance designed for a company a hundred times its size.

The experiment is celebrated as evidence of openness while being denied the conditions required to succeed.

When it struggles, the organisation learns exactly the wrong lesson. The market was not ready. Customers did not understand. The economics were unproven. The start-up lacked discipline. The old model remains safer.

The future is permitted to fail only in forms which leave the present unthreatened.

Cannibalisation sits at the centre of the trap. Companies want new revenue without damaging old revenue; simpler products without reducing prices; automation without removing profitable work; and new channels without weakening established distribution.

Real change is rarely so accommodating.

An incumbent may need to damage a successful business before somebody else does it more thoroughly. The damage is immediate and measurable, while the benefit is speculative and distant. Lost revenue appears this quarter. Strategic value may take years to emerge. The executive responsible for the sacrifice may never remain long enough to receive credit.

Waiting is easier to defend.

Waiting preserves the numbers, avoids embarrassing write-offs and gives the current strategy another chance to work. Every delay feels prudent until acting requires changes so severe that the organisation can no longer tolerate them.

By then, the problem has become entangled with identity.

Large organisations build cultures, hierarchies and narratives around their products. Leaders become important because they understand them. Teams gain status because they control them. Investors value the company through them. Entire careers depend on the belief that a particular market, technology or operating model remains important.

Accepting that the basis of value has changed can mean accepting that much of this expertise no longer provides the advantage everybody believed it did.

The organisation responds by redescribing its liabilities as strengths. Complexity becomes sophistication. Switching friction becomes integration depth. Slow governance becomes enterprise readiness. Comprehensiveness becomes strategic breadth.

The people authorised to change the system are usually those whose authority was created by it.

I’ve described the same mechanism at the level of technology. Technical debt becomes an adaptation problem when every change requires certainty, coordination and months of work; and architecture becomes a political structure when systems, ownership and authority harden around one another.

The company is trapped between two machines. One has customers, revenue, operating history, political support and measurable returns. The other has assumptions. The first can prove its value, so it wins whenever resources tighten.

The incumbent therefore has enormous capacity in theory and remarkably little freedom in practice. Its resources are tied to obligations, its talent is organised around existing products, and every consequential decision must survive contact with the organisation it is intended to transform. It can fund pilots but not transitions; buy options but not commit.

Transformation is not rejected. It is domesticated.

Historically, incumbency gave companies years to remain caught between recognition and response. Their size absorbed mistakes, their brands reassured customers and their contracts continued to renew.

That protection is becoming less dependable. The machinery required to challenge them is getting cheaper, and valuable parts of a market can move before the loss becomes visible in aggregate.

Zombie companies have always existed. What is changing is how long they can remain convincingly alive.

The cost of challenging incumbents is falling

For most of modern business history, incumbency bought time.

A challenger could identify an obvious weakness, understand exactly why customers were dissatisfied, and still spend years trying to build something credible enough to matter. It needed capital, infrastructure, specialist expertise, distribution, customer support, regulatory knowledge, sales capacity and enough trust to persuade buyers that choosing the smaller company would not become a career-limiting mistake. The incumbent could be slow, because everybody else had to be slow too.

That gave large companies a generous margin for error. They could misread emerging behaviour, dismiss early competitors, postpone difficult decisions and recover after the evidence became clear. Even when the diagnosis was obvious, assembling the machinery required to exploit it remained expensive.

A terrible product could survive because building a better one was hard. A hated supplier could survive because replacing it was harder. An inefficient intermediary could survive because customers lacked the knowledge, access or confidence to navigate the market without it. Some of those advantages are genuine moats. Others merely reflect how expensive competition used to be. Those costs are falling.

A small company can rent infrastructure which once required enormous capital investment. It can access global distribution through existing platforms, assemble products from mature services, outsource specialist functions and reach customers without first building a large sales or marketing organisation. Open-source software, cloud infrastructure, online communities, payment systems, logistics networks and global talent markets have removed much of the institutional mass previously required to produce a credible alternative.

The challenger still needs to build something useful. It still needs judgement, focus, trust and a reason to exist. But it no longer needs to become a miniature version of the incumbent before it can compete. That changes the shape of the threat.

Large organisations tend to imagine disruption as replacement. They picture a rival growing until it can serve the same customers, provide the same breadth, satisfy the same regulations and operate the same complex business at comparable scale. That is reassuring because replication is difficult.

The incumbent knows how much machinery sits behind the polished surface of its product. It knows how many people, systems, contracts and compromises are required to keep everything running. When a new entrant appears with a simpler proposition and a fraction of the workforce, it often looks naive. Perhaps it is. Or perhaps it has correctly identified that customers only value a small part of the machinery.

The challenger does not need to conquer the entire market. It can take one profitable workflow, one frustrated customer group, one emerging channel, one awkward use case or one stage of the value chain where the incumbent’s complexity creates more friction than value. It can ignore everything else.

This is particularly dangerous because large companies tend to subsidise their least attractive activities with their most attractive ones. Profitable customers support expensive infrastructure. Simple transactions fund complex service obligations. High-margin products carry weaker ones. Bundles hide the economics of individual components.

A focused competitor can extract the valuable part without inheriting the structure surrounding it.

The incumbent keeps the legacy systems, the difficult customers, the regulatory obligations, the low-margin support burden and the promises made under an earlier version of the market. The challenger takes the clean edge.

This rarely produces an immediate collapse. In fact, the first effects can make the incumbent appear stronger. As smaller or less profitable customers leave, average revenue per customer may rise. As the company retreats towards its largest accounts, margins may improve. As weaker competitors disappear, market share may increase. As products are consolidated and prices rise, the business may look more disciplined.

The organisation reports better numbers, while the quality of the underlying opportunity deteriorates. This is how hollowing out hides inside an apparently sensible strategy.

The company focuses on its best customers, only to discover later that those customers were once smaller, less sophisticated customers whom the market used to funnel towards it. It concentrates on the most profitable workflows, while new platforms redefine which workflows matter. It protects its enterprise position while losing the environments where tomorrow’s enterprises are being formed. 

The revenue remains. The replenishment mechanism disappears. But by the time leadership notices that the pipeline contains fewer genuinely new customers, the competitor may no longer look like a competitor. It may be a platform, a workflow, an interface, a marketplace or a behaviour which routes around the old category altogether.

That is another assumption incumbents often get wrong. They expect the thing which displaces them to resemble them.

A traditional agency looks for another agency. A software vendor looks for another software vendor. A publisher looks for another publisher. A bank looks for another bank. Meanwhile, the customer solves the problem through a collection of tools, platforms and services which individually appear too narrow to count as a direct threat.

The category is not conquered. It is disassembled.

One company absorbs discovery. Another owns the transaction. Another automates the specialist work. A platform controls the customer relationship. A lightweight tool handles the most common use case. The remaining complexity is pushed towards the incumbent, which continues to describe its comprehensiveness as a strength.

The incumbent may still have the best complete solution, but completeness becomes less valuable when customers no longer want a complete solution.

As I argued in The middle is a graveyard, better comparison weakens the value of breadth and acceptable competence. A small product may have little revenue but enormous influence over expectations, training customers to expect a simpler interaction or a different price. The market often changes culturally before it changes financially.

None of this means that the challenger is necessarily better. Many challengers will fail. Some will discover that the complexity they removed was carrying hidden value. Others will become bureaucratic as soon as scale introduces customers, obligations and politics of their own.

But the market does not need every challenger to succeed. It needs enough of them to keep testing which parts of inherited advantage remain defensible.

That is the new pressure. Zombie companies once benefited from the enormous cost of proving that they were no longer necessary. Now, different parts of the market can run that experiment simultaneously.

One competitor tests whether the product can be cheaper. Another tests whether it can be simpler. Another tests whether the service can be automated. Another tests whether the category can be bundled into a broader platform. Another tests whether customers need the underlying activity at all. The incumbent must defend every valuable edge. The market needs only enough experiments to make one of those edges contestable.

This does not mean that scale has stopped mattering. Capital, distribution, trust, data, regulation and operational competence remain formidable advantages. A large organisation can absorb shocks which destroy smaller competitors, and many incumbents will use their resources to adapt successfully.

But scale no longer provides the same guarantee that challengers must first reproduce the incumbent’s institution before they can threaten its economics. They can rent the capabilities they need, avoid the obligations they do not, and enter through whichever weakness offers the clearest path.

The consequence is a shortening gap between strategic failure and commercial recognition.

A company can still coast, but the runway is becoming shorter. Customers learn more quickly. New products improve faster. Alternatives spread through communities and workplaces. Attractive parts of the market can move before the incumbent sees a dramatic change in total revenue. The protective delay remains, but it has become less dependable. 

And there is another force compressing it further.

The costs of building alternatives are falling, but so are the costs of understanding why the incumbent is weak, comparing the available options and turning dissatisfaction into action. The market is gaining better tools for inspecting companies, evaluating trade-offs and navigating away from products which no longer deserve their position.

AI matters here, although not for the reason most corporate strategy decks suggest. Its most dangerous effect may be that it makes failure easier to see.

AI shortens the period in which failure can hide

The corporate conversation about AI is mostly obsessed with production.

Can we write more quickly? Build software with fewer people? Automate support? Reduce headcount? Generate campaigns, presentations, reports, images and strategies without paying quite so many humans to move rectangles around documents?

Those questions matter, particularly for companies whose costs depend heavily on knowledge work. But they frame AI as a tool which businesses use internally, while overlooking the more dangerous change happening outside them.

The market is gaining the ability to understand companies more quickly.

That sounds less dramatic than replacing entire departments, but it attacks one of the conditions which allows zombie organisations to survive: the gap between strategic decay and everybody else noticing.

For decades, large companies benefited from the cost of investigation. Customers could see the product, the price, the marketing and perhaps a handful of reviews, but understanding the whole organisation was prohibitively difficult. The reality was scattered across documentation, support threads, employee experiences, technical decisions, pricing structures, regulatory filings, customer complaints, product updates and thousands of small interactions which nobody had the time or expertise to assemble.

A recognisable brand compressed that complexity into a reassuring shortcut.

The customer did not know whether the company was still competent, responsive or particularly good. They knew that it was large, familiar, widely used and unlikely to disappear next Tuesday. In a market where proper due diligence was expensive, those signals carried enormous value.

That protective ambiguity is weakening. As I argued in When scrutiny becomes cheap, AI does not need to become omniscient to change the balance. It only needs to make questions which were previously too expensive to ask cheap enough to ask for that to become routine.

The answers will be imperfect. The evidence will be noisy. Models will misunderstand context, amplify weak sources, and occasionally invent things. But perfection is unnecessary. Many incumbents did not need buyers to reach perfect conclusions. They benefited from buyers being unable to investigate at all.

That’s the important economic change. When inspection becomes easier, scale becomes a weaker proxy for competence. Familiarity carries less reassurance when the machinery behind it can be examined. The brand still shapes the question, but it no longer controls the evidence available to answer it.

And scrutiny is only the beginning. In Marketing after the fog clears, I explored how cheaper comparison exposes the trade-offs which marketing once softened or concealed. Every company has limitations, exclusions and compromises. AI makes it easier to find them, but also to determine whether they are intentional, justified and appropriate for a particular buyer.

That changes the value of incumbency. The largest company often benefited from being the safest answer to an imprecise question. AI can make that question specific to a team’s budget, systems, constraints, risks and tolerance for complexity.

The incumbent may still be the right answer. Frequently, it will be. Scale can buy reliability, depth, support and resilience which smaller competitors cannot provide. But it must win the comparison rather than merely embody safety. A narrower competitor can be recommended for the particular job without proving that it is better in every possible dimension.

That precision also accelerates unbundling. An assistant can identify that one tool handles the common workflow more simply, another provides the occasional specialist capability, and the expensive incumbent mostly survives because buying everything together used to be easier than evaluating the pieces.

The comprehensive suite becomes a set of individually contestable decisions.

That does not necessarily cause mass cancellation. Contracts, integrations and internal expertise continue to create friction. But the next purchase becomes less automatic. The next team chooses differently. The next generation of users never develops the dependency which made the incumbent difficult to displace. Recommendation changes behaviour before churn changes revenue.

As I argued in The inversion of purpose, AI systems increasingly sit between intention and action. They will not be neutral. They may reproduce brand bias, favour familiar vendors, reward whatever is best documented, or simply get the answer wrong.

But they create another evaluation surface which the incumbent does not fully control. A company may perfect its funnel and discover that the shortlist was assembled before the customer arrived. Incumbency still matters, but it must survive a new layer of interpretation rather than merely dominate the old one.

AI also reduces the cost of acting on dissatisfaction.

Knowing that a product is poor has never guaranteed that customers will leave it. Migration requires planning, documentation, retraining, data mapping, integration work, internal persuasion and the careful management of all the obscure dependencies discovered shortly after somebody announces that the transition should be straightforward.

Those costs have protected bad products for decades.

AI will not make complex migrations effortless, but it can make parts of them ordinary. It can document existing workflows, translate formats, map fields, generate scripts, identify dependencies, compare contracts, produce internal guidance and help somebody construct a credible business case for leaving.

Each reduction in friction makes dissatisfaction slightly more dangerous.

This creates a feedback loop which zombie companies are poorly equipped to recognise. Easier inspection reveals more weakness. Easier comparison gives that weakness commercial context. Easier action makes the conclusion consequential. As more customers experiment with alternatives, more evidence appears for the next customer and the next recommendation system to evaluate.

The incumbent’s accumulated reputation once dampened uncertainty. Now its accumulated residue can compound against it.

AI does not decide that a company is dead and announce the result to the market. The process is quieter and more distributed than that. Thousands of individual questions become slightly easier to ask. Thousands of buyers receive slightly more specific comparisons. Thousands of teams find it slightly easier to try, justify or migrate towards something else. 

The aggregate effect arrives later, in the metrics the company trusts. That is what makes this so dangerous. AI compresses the period between an organisation becoming strategically unfit and the market becoming capable of behaving as though it knows.

The zombie still has customers, revenue and a category-leading product. But it has less control over how those facts are interpreted, less protection from the evidence surrounding them, and less time to respond before changing perceptions become changing behaviour.

You can see the shape of that problem clearly in mature software markets, where an incumbent may continue dominating the product category while cheaper tools, simpler workflows and changing user needs gradually make the category itself less central.

Winning the wrong market

Consider Photoshop. This is an illustration rather than a prediction about Adobe, but it makes the mechanism easier to see.

For decades, professional image editing required specialist software, specialist knowledge and enough commitment to learn an interface built around layers, masks, channels and colour spaces. Adobe built an extraordinarily powerful product around that reality, then surrounded it with training, file formats, professional habits and organisational workflows which made Photoshop increasingly difficult to avoid.

Those advantages remain formidable. Entire industries depend on the product, and a competitor cannot reproduce a feature list and expect decades of embedded behaviour to disappear.

But Photoshop’s future depends on more than whether somebody can build a better Photoshop.

Canva changed who could produce competent visual material by turning much of the work into templates and constrained choices. Social platforms changed what gets made, where it appears and what counts as good enough. Mobile tools moved parts of the workflow away from the desktop. Generative systems increasingly allow people to describe an outcome without learning the mechanics traditionally required to produce it.

None of these needs to replace Photoshop completely. They only need to remove reasons for people to become Photoshop users in the first place.

A marketing manager can adapt a template. A small business can generate a passable campaign without building a creative department. A creator can produce and publish without leaving the platform where the audience already lives. Somebody who needs to remove a background or resize an image can use a narrow browser tool without learning a professional suite.

The market fractures into outcomes.

Photoshop may remain the most powerful answer for professional image editing while professional image editing becomes a smaller part of how visual work gets done. It can continue winning within its category as more activity moves into simpler tools, automated workflows and platforms which barely present themselves as image editors.

That is the strategic danger conventional competitive analysis struggles to capture. Incumbents look for products which resemble their own. They compare feature lists, enterprise adoption, professional credibility and market share. Those are sensible questions while the category remains stable; they become distracting when customers are beginning to bypass it.

The consequential competitor may lack most of the incumbent’s capabilities because most customers no longer need those capabilities. Its apparent weakness is the source of its appeal. It removes decisions, hides complexity and delivers an acceptable result before the incumbent has finished presenting its options.

A professional sees limitations. A customer sees that the job is done.

Depth, control and expertise still matter. Professional users still need powerful tools, and large organisations still require reliability, governance and compatibility. Adobe may navigate the transition brilliantly.

The risk sits elsewhere. The economics of a category-leading product may depend on a much wider population than the specialists who genuinely need its full capability. It needs students to learn it, small companies to adopt it, junior employees to bring it into organisations and occasional users to tolerate a comprehensive tool for a narrow task.

When those pathways weaken, the installed base can remain enormous while its replenishment mechanism decays.

The business may even appear to improve. Professional customers remain embedded. Average revenue rises. Low-value support costs fall. The company concentrates on users who appreciate its depth and can afford its pricing.

That may become a good, profitable and durable business. It is simply a different business from the one implied by category dominance.

Market share can tell Adobe how much of professional image editing it controls. It cannot tell it how much visual work has escaped professional image editing entirely.

Dominance is only valuable while the thing you dominate still matters.

That is the problem with measuring strategic decay from inside the category being eroded.

Dashboards inherit the assumptions

The obvious response is to measure more carefully.

Track acquisition by cohort. Monitor emerging competitors. Analyse changing behaviour. Build leading indicators. Commission research. Add another dashboard, preferably one with an ominous name and several arrows pointing towards the future.

Large companies are rarely short of data. Every interaction generates an event, every team has objectives, and every executive has a scorecard.

But measurement systems inherit the assumptions of the organisations which create them. As I argued in Clicks don’t count, the metrics we can observe often describe the interface rather than the underlying competitiveness.

A subscription business becomes exceptionally good at measuring subscriptions. A retailer understands transactions. An enterprise software company tracks seats, renewals, expansion revenue and product adoption. Those metrics can be accurate and operationally essential while remaining blind to the possibility that the machinery itself is becoming the problem.

A dashboard can show that a customer renewed. It cannot easily distinguish affection from inertia, contractual timing, migration risk or the exhausted resignation of somebody who cannot face another procurement process. It can show that usage remains high while missing that the activity has become an obligation. It can show that a feature is widely adopted while missing that the feature compensates for a workflow customers would rather abandon.

The metric is correct. The interpretation is convenient.

High retention becomes loyalty. High usage becomes engagement. Switching costs become integration depth. Price tolerance becomes willingness to pay. Each translation turns a historical advantage into a reassuring statement about current fitness.

Measurement also becomes political. Every metric has an owner, an audience and a consequence. Product needs adoption to rise. Sales needs pipeline. Finance needs margin. A transformation programme needs milestones. The chief executive needs a coherent story for the board.

The organisation negotiates which version of performance becomes official.

Signals which support the current model are quantifiable and easy to defend. Signals which challenge it arrive with caveats. A competitor is gaining cultural relevance, but its revenue remains small. New users prefer a different workflow, but they are not valuable customers yet. Satisfaction is declining, but renewal remains strong. The category is fragmenting, but the company still leads every report built around its old boundaries.

These signals create arguments. The dashboard offers a cleaner reality.

Evidence about the present is treated as fact. Evidence about the future is treated as opinion.

The asymmetry extends into decision-making. Doing something new requires forecasts, validation, financial modelling and an explanation of how the investment will produce returns. Continuing the existing strategy requires only the absence of a sufficiently dramatic reason to stop.

The burden of proof belongs to the future.

Aggregation compounds the problem. Large businesses need averages and totals to remain governable, but averages are where emerging futures disappear. A collapse among younger users can be outweighed by expansion among older accounts. A weakening entry-level market vanishes beneath enterprise renewals. The declining past remains much larger than the growing future until, suddenly, it does not.

Even sophisticated segmentation reflects the organisation’s existing structure. It compares enterprise with small business, region with region and product tier with product tier.

The market may be reorganising itself around customers who want outcomes rather than tools, teams assembling lightweight workflows, businesses which never adopt the supposedly essential system, or buyers whose shortlist was assembled by an assistant.

Those behaviours cut across the dashboard because no department owns the question which would reveal them.

The same blind spot affects competitive intelligence. Companies are excellent at monitoring recognised rivals. The threat may instead be a platform absorbing the customer relationship, a workflow assembled from several narrow tools, or a behavioural change which eliminates the purchase.

Dashboards struggle with disappearance. They cannot easily count the customer who never entered the funnel, the task which no longer required the product, or the decision which moved beyond the category.

Absence produces very little data.

The organisation sees fewer opportunities and explains them through execution. Marketing must create more demand. Sales needs better enablement. Product needs stronger differentiation. The brand needs refreshing.

Each response assumes that the market still contains the same decision and that the company is merely performing poorly within it.

Sometimes the decision has moved.

Weak signals could reveal that movement, but organisations process them into administration. Customer frustration becomes a product issue. The product issue becomes a roadmap consideration. The roadmap consideration becomes one of several priorities. By the time it reaches leadership, the warning has become manageable, owned and reassuringly scheduled.

A risk register may contain every ingredient of the company’s eventual failure, each assigned to a different owner with a mitigation plan. Nobody owns the possibility that the collection of manageable risks describes an unmanageable trajectory.

More metrics cannot solve that problem.

A company can measure emerging relevance, adaptability, experiments, new cohorts and revenue from new products. Those measures help only when the organisation is prepared to accept what they imply.

If declining relevance among new users can never outweigh revenue from existing ones, the measure is decorative. If evidence that customers are trapped does not change how retention is interpreted, the research is theatre.

Zombie companies do not necessarily lack information. They lack a decision-making system capable of privileging future viability over present reassurance.

The numbers can expose the trade-off. They cannot make the organisation choose.

Eventually, the choice disappears. What remains is the accounting.

The accounting lag

Companies rarely fail when the share price falls, the customers leave, or the redundancies begin. But those are the moments when the failure becomes visible.

The real failure happened earlier, when the organisation stopped being able to create a credible future for itself. It happened when protecting the current product became more important than understanding the changing problem. When new customers began forming habits elsewhere. When every proposed alternative was required to preserve the economics, politics and status of the existing business. When the company could still see several possible futures, but could no longer move towards any of them.

Everything after that is just delay.

The brand keeps generating demand. Contracts continue to renew. Switching remains painful. Sales teams keep harvesting relationships. Prices rise. Costs fall. Acquisitions provide fresh stories. Another reorganisation creates the appearance that the company has finally understood the urgency.

The organisation remains busy, confident and financially productive while the available future quietly narrows around it.

Eventually, the market catches up. A competitor becomes credible. A workflow moves elsewhere. A platform absorbs the relationship. A generation of customers arrives without the assumptions which made the incumbent feel essential. AI makes the weaknesses easier to inspect, the alternatives easier to compare and the cost of leaving easier to tolerate.

The financial consequences can then appear suddenly, because finance records the crossing of thresholds rather than the years spent approaching them.

Revenue slows. Acquisition costs rise. Discounts increase. Customers reduce their commitments. The company announces that demand has softened, conditions have changed, or execution has not met expectations. Leadership promises greater focus. A strategic review begins.

By then, the market is reporting old news. The company did not fail when customers left, the category declined or a new technology arrived. Those events merely exposed that it had spent years protecting a model it could no longer escape.

The collapse, when it comes, is an accounting event. The strategic event happened years earlier.

Not every ageing incumbent becomes a zombie. Accumulated advantage can provide extraordinary freedom. Brand, capital, distribution, data and expertise can widen the range of futures available to an organisation, allowing it to make bets which smaller competitors could never survive.

But those advantages only remain alive when they are converted into new capability. Otherwise, they just become reserves.

The company spends them slowly, mistaking the continued ability to pay for things as evidence that it still knows what to build. It uses its strength to postpone choosing until the market removes the choice entirely.

That is the line between resilience and undeath.

A living organisation can still make meaningful decisions about what it becomes next.

A zombie can only continue.

And by the time the numbers show the difference, the difference will have been true for years.

Label

Name*

Email*

Label

Name*

Email*

0 Comments