Successful companies can become highly efficient at the business they already know while steadily reducing their capacity to question it. Decades of management research suggest that repeated profits can favour familiar investments, narrow the evidence reaching executives and make retreat from earlier decisions more costly.
This does not mean success inevitably produces complacency. Nor does a failed investment prove that the original decision was foolish. Companies make choices under uncertainty, and many promising technologies and business models never become commercially viable.
The problem is more specific. Success rewards certain routines, people and assumptions. As those choices become embedded in budgets, skills and reporting systems, the company gets better at repeating them and may become less capable of testing alternatives.
Why success favours what already works
James G. March gave management researchers a useful way to describe this tension in his 1991 Organization Science paper on exploration and exploitation.
Exploration means searching, experimenting, taking risks and testing unfamiliar possibilities. Exploitation means refining and extending knowledge the organization already possesses.
Companies need both. Exploitation keeps factories productive, improves established products and turns experience into lower costs or better service. Exploration creates the possibility of finding a different technology, market or business model.
However, their rewards arrive on different schedules. An improvement to an existing process can produce a measurable return this quarter. A new technology may consume money for years and still fail. When managers are evaluated on near-term results, familiar projects often have the stronger case.
March used formal models rather than a representative survey of companies, so his paper did not establish that every successful business eventually becomes trapped. Its contribution was to show how adaptive learning can favour exploitation in the short run even when too little exploration creates a longer-term weakness.
Daniel Levinthal and James March later called one version of this problem the “success trap”. Organizations learn most easily from results that are recent, nearby and connected to activities they already perform. Distant possibilities and failures that never reached the market are harder to observe.
Success changes the evidence a company sees
There is an easily overlooked implication here: a company’s strategy helps determine which evidence the company will later use to judge that strategy.
If most investment goes into an established product, management receives a continuous stream of information about that product. It sees customer orders, margins, quality problems and the effects of each incremental improvement. An unfamiliar alternative, funded by a small experimental budget, produces much less information and usually carries wider error margins.
The mature business can therefore appear to have stronger evidence behind it partly because the company has spent years generating that evidence. This does not make the figures false. It means the comparison is uneven.
Profits from the existing operation answer one question: does this activity still produce money? They do not automatically answer a different question: is this the best place for the next pound or dollar of investment?
That distinction matters when technology, regulation or customer behaviour changes. A company can make a perfectly accurate forecast for its current market while failing to study a smaller market that is developing under different economics.
Executive confidence can affect investment
Success also affects the people allocating the money. Senior executives usually reach their positions after a series of achievements, giving them genuine reasons to trust their judgement. Confidence is necessary when decisions must be made before all the facts are available. Overconfidence is different because it can lead managers to overestimate likely returns or the accuracy of their forecasts.
Ulrike Malmendier and Geoffrey Tate examined this issue in a 2005 study of CEO overconfidence and corporate investment. Their data covered the personal portfolio decisions and company investments of chief executives at 477 large US companies between 1980 and 1994.
The researchers used behaviours such as repeatedly holding company share options longer than diversification theory would normally predict as proxies for overconfidence. These were indirect measures, not psychological diagnoses.
They found that investment under the CEOs classified as overconfident was more sensitive to internally generated cash, particularly at companies that depended more heavily on equity financing. The authors’ interpretation was that these executives could overestimate returns from their own projects while believing outside investors undervalued the company.
The study was observational. It identified relationships in a historical sample of large US businesses and did not prove that corporate success caused an executive to become overconfident. Even so, it showed why company accounts alone may not fully explain investment decisions. The judgement of the person reading those accounts matters too.
Sunk costs make withdrawal harder
Once a company has committed money and reputation to a project, another mechanism appears. Managers may continue investing because stopping would acknowledge that earlier expenditure cannot be recovered.
Economists call that irrecoverable expenditure a sunk cost. It should not determine whether the next investment is worthwhile. The relevant comparison is between the expected future costs and benefits of continuing from today, not the amount already spent.
Barry Staw demonstrated the basic problem in a 1976 simulated business investment experiment. Participants who were responsible for an earlier allocation could commit more resources after receiving negative feedback than people who had not made the original decision.
The wider literature calls this escalation of commitment. A later multilevel review of the research found that escalation cannot be reduced to a single cause. Group relationships, organizational conditions and forces outside the company can all affect whether commitment continues.
This helps explain why a weak project can survive several reviews. Each additional investment increases the loss that withdrawal would expose, while any remaining possibility of recovery gives supporters a reason to ask for more time.
“Groupthink” is not a complete diagnosis
Poor corporate decisions are also frequently attributed to groupthink, the idea associated with psychologist Irving Janis that pressure for agreement can stop a cohesive group from examining alternatives properly.
The term is useful shorthand, but it is often applied too casually. A review of 25 years of groupthink research found a relatively small empirical literature and continuing theoretical and methodological problems. Glen Whyte also argued that some apparent groupthink cases could be explained through framing, risk-seeking after losses and group polarization.
It is therefore too simple to assume that a close management team will inevitably silence dissent. The more useful questions concern the actual decision process. Who controls the agenda? Which evidence reaches the meeting? Is the project sponsor also judging whether the project should continue? What happens to a manager who challenges the chief executive?
Ten experienced people around a table do not provide ten independent assessments if all of them are working from the same assumptions and incentives.
Structure preserves yesterday’s priorities
Strategy eventually becomes physical. A company builds plants, signs supplier agreements, trains employees and creates performance measures around its chosen activities. Departments acquire budgets and authority. Managers develop careers based on particular products or methods.
These arrangements are not simply bureaucratic obstacles. Standard routines can make a company reliable and accountable. However, organizational theorists Michael Hannan and John Freeman argued that the same qualities are connected to structural inertia, which makes major change more difficult.
A new strategy may therefore threaten more than current revenue. It may reduce the usefulness of existing assets, demand unfamiliar skills and shift influence away from the people who built the old business. Saying “change direction” is much easier than changing those connected systems.
How companies can make reconsideration easier
There is no checklist that can remove uncertainty or guarantee sound decisions. Management research does, however, point towards processes that address the mechanisms described above.
Charles O’Reilly and Michael Tushman use the term organizational ambidexterity for the capacity to exploit mature activities while also exploring new technologies and markets. The two sides require different conditions. Established operations benefit from consistency and cost control, while experiments need room for uncertainty and failure.
In practical terms, companies can protect a defined amount of money and management attention for exploration instead of expecting experimental projects to compete immediately with mature operations. They can specify performance milestones and withdrawal conditions before a project begins, when nobody’s reputation is yet tied to defending it.
A review can also be led by someone other than the executive who originally sponsored the investment. Assumptions recorded at approval can be compared with current evidence, making it harder to rewrite the original justification after results disappoint.
These measures are safeguards, not proven universal cures. Exploration can waste money, independent reviewers can be wrong and a project that misses an early target may still succeed. The purpose is to make changing course a normal management decision rather than a personal confession of failure.
Success gives a company resources, expertise and detailed knowledge of its customers. It also gives the company routines that generate more evidence in favour of what it already does. The difficult question for management is whether strong results from today’s business also justify making the same investment tomorrow.
