CEO perspective
In a recent conversation, a sentence came up that I have heard so many times – across the telecom industry, within tech teams, even in everyday conversations around the dinner table:
“We can’t change direction now. We’ve already invested too much.”
You hear it when someone keeps repairing a car that should have been replaced two invoices ago. You hear it during a house renovation when the budget has doubled but the logic feels irreversible. And you hear it in business when teams keep building a platform – even when deep down everyone knows the market has moved, the assumptions have changed, or the original plan is no longer the rational choice.
This is neither irrational nor unusual. It is a very human decision pattern with a name: the sunk cost fallacy – closely related to what organizational psychologists call escalation of commitment.
In rational decision-making, past spend is past. Only future costs, future benefits, and opportunity costs should guide the next decision. But organizations are made of people, not spreadsheets – so the “rear-view mirror” often wins.
In the last CEO Blog, I focused on a simple idea: cheaper is always an option – but what are you really choosing? The real challenge begins after the decision is made. Because even after you choose a path – cheap or expensive, build or buy, internal or external – the most dangerous cost is the one that traps your future decisions.
What the sunk cost fallacy really is
The sunk cost fallacy is not simply “spending too much.” It is the tendency to continue an endeavor because we have already invested money, time, or effort, even when evidence suggests continuing is unlikely to be the best option.
This is not a new discovery. Organizational psychologists have been studying this behavior for more than 50 years. In the research literature, escalation of commitment shows up most clearly in repeated decisions under negative feedback: a project underperforms, but instead of cutting losses, the decision-maker commits more resources and takes on more risk.
Barry Staw’s early work already demonstrated a key accelerator: when people feel personally responsible for the earlier decision, they are more likely to “double down.” Decades of research since then have only reinforced the same core drivers- particularly responsibility for the initial decision, sunk costs, and proximity to completion (the powerful feeling that “we’re too close to stop now”).
What matters for leadership is this: the sunk cost fallacy is not about intelligence. It is about predictable human psychology that becomes stronger under pressure – especially organizational pressure.
The psychological drivers that make smart leaders stay on the wrong path
The “why” behind sunk costs is not one single bias. It is a stack of forces that reinforce each other.
Loss aversion and the pain of “making the loss real”
Prospect Theory describes a basic asymmetry: people experience losses as more psychologically powerful than equivalent gains. In many empirical estimates, the “loss” side of the value function is roughly twice as steep as the “gain” side – often summarized as “losses loom larger than gains.”
Translated into project decisions: stopping a project feels like locking in a loss, while continuing preserves the hope – however unrealistic – that the loss can be avoided.
Cognitive dissonance and self-justification
Cognitive dissonance theory explains how inconsistency between “I am competent” and “I made a bad bet” creates discomfort – and how people reduce discomfort by changing their interpretation of reality, rather than changing course.
In organizations, this often becomes self-justification: we reinterpret signals (“the turnaround is almost here,” “the data is noisy,” “the market will come back”) so that continuing feels consistent with competence. Major escalation research and reviews repeatedly connect escalation to self-justification dynamics.
Reputation pressure and “trusting the stubborn leader”
Here is the uncomfortable part: escalation can be socially rewarded. Research on the reputational causes and consequences of escalating commitment shows that observers can trust decision-makers who escalate, which helps explain why escalation persists even when it is economically harmful.
In leadership roles, this becomes a trap: if persistence is culturally coded as strength, then a rational stop decision can look like weakness, even when it is the best move for the company.
The “don’t waste” rule
A powerful – and often moral – driver is the internalized rule “don’t waste.” Research on sunk costs highlights that the bias can be fueled by an overgeneralization of that rule: because waste feels wrong, people keep investing to avoid the feeling that earlier resources were “wasted.”
In business, this moral framing is dangerous because it replaces a forward-looking question (“what creates value from here?”) with a backward-looking emotion (“how do we justify what we already spent?”).
How this shows up in messaging and telco
Telecom and messaging are unusually exposed to sunk-cost thinking – not because the industry is irrational, but because the environment amplifies the triggers.
High sunk investments, long cycles, and public accountability
Telecom operators face constant pressure to monetize massive infrastructure investments, and the role of IT and technology capabilities becomes central to performance.
When investments are large and visible – networks, platforms, migrations – decision-makers carry not only financial responsibility but reputational and political responsibility. That creates ideal conditions for escalation.
Messaging’s unique tension: “mission-critical” yet constantly shifting
Messaging looks simple from the outside, but it has become mission-critical infrastructure for operators, enterprises, and service providers – and it evolves continuously under pressure (fraud, compliance requirements, new channels, changing enterprise expectations).
That mix – pressure plus change – creates the perfect environment for sunk cost traps. Three common patterns appear in our sector:
Build traps in platforms and integrations
Teams start building “our own platform” because it seems cheaper, faster, or more controllable – until the build hits reality: edge cases, throughput, redundancy, routing complexity, security, compliance, and ongoing maintenance. At that point, switching to a proven solution feels psychologically impossible because it would mean admitting that earlier effort is not recoverable. This is a textbook sunk cost pattern (past effort driving future commitment).
Security modernization traps
The fraud landscape in business messaging evolves quickly. For example, the Mobile Ecosystem Forum report on A2P SMS monetisation describes how first-generation SMS firewalls were designed to block spam, but today the threat set has multiplied and older approaches can become ineffective.
A common organizational risk (and this is an inference from the psychology, not a claim about a specific operator) is continuing to “patch what we have” mainly because we already paid for it – rather than evaluating whether the next euro is better spent on the capability that actually protects trust and revenue now.
Strategy traps in A2P monetisation
GSMA Intelligence notes that with P2P SMS cannibalised by OTT services, A2P SMS represents the greatest messaging revenue opportunity for mobile operators – while also facing headwinds that require an active strategy.
That combination – high revenue potential plus headwinds – can create escalation behavior: organizations keep investing in the same commercial model, the same routing logic, or the same partner structure because they have already built it, instead of revisiting whether the next investment should change the model.
Historical mega-projects simply make the pattern visible
The point is not that “telecom is like an airport project.” The point is that human escalation patterns repeat across domains. Berlin Brandenburg Airport opened roughly a decade late and at around €7 billion versus early estimates around €2 billion.
And the “Concorde effect” is so iconic that it became shorthand in psychology for sunk-cost persistence.
These are not “stupid decisions.” They are what happens when sunk costs, reputation, and self-justification become stronger than forward-looking economics.
A CEO-grade de-escalation playbook
Knowing the bias exists is not enough. Escalation of commitment persists even when people can name it.
The solution is not “be more rational.” The solution is to build decision systems that make rationality easier.
Run a “start today” test, not a “finish what we started” test
Ask: If we were starting today – with today’s information – would we choose this path again?
This aligns with the economic logic that sunk costs should not be decisive inputs, and it short-circuits self-justification.
Pre-commit to exit criteria
Escalation thrives when teams can always say “just a bit more.” Instead, define kill criteria early: measurable thresholds for value, timeline, and risk that trigger a stop/pivot decision. Project-management literature explicitly frames escalation as continuing despite information that expected return will not be achieved – so the antidote is to define what “not being achieved” means before emotions take over.
Use a premortem to surface the truth people are afraid to say
The “premortem” method asks teams to assume the project failed and then list reasons why – creating a structured way for concerns to surface before commitment escalates further.
Separate identity from decisions and force an “outside view”
When decision-makers feel personally responsible for the initial choice, escalation increases.
This is why governance matters: independent review, red teams, and decision forums where changing course is treated as competence, not failure. Inside views turn unique project details into overconfidence, while the outside view offers a corrective by comparing the initiative to similar efforts and grounding expectations in real-world outcomes, helping reduce bias and support more objective, rational decision-making.
In telco environments, nearly 70% of digital transformation initiatives have failed to achieve their intended outcomes over the past two decades. This track record of underperformance has fostered hesitation, skepticism, and stalled business cases. When this “history of failure” becomes a justification for repeating the same approaches, it creates a second-order sunk cost problem – where past losses continue to distort future decisions.
What this means for HORISEN and for our customers
At HORISEN, we talk a lot about long-term thinking, consistency, and independence – not as slogans, but as operational discipline.
The sunk cost fallacy matters for us on two levels.
What this means for HORISEN as a product company
We must be willing to do what sunk-cost psychology resists: rebuild, refine, and rethink platforms when necessary – not because change is fashionable, but because reliability, scale, and relevance demand it.
This also reinforces why we build modularly. A modular ecosystem is not only a technical preference – it is an organizational bias-corrector. Modularity creates exit paths. It makes “swap, replace, evolve” feasible without catastrophic rewrites, which reduces the psychological pressure to defend one monolithic bet forever.
What this means for messaging operators, aggregators, and enterprises
Our customers face the most punishing version of sunk costs: once messaging infrastructure is deployed, it becomes embedded in routing, billing, compliance, customer SLAs, fraud controls, and enterprise commitments.
That is exactly why vendor-neutrality and integration readiness matter: they preserve strategic freedom.
In practical terms, resisting the sunk cost fallacy in messaging often looks like:
Choosing architectures that keep replacement possible
This is not “build vs buy” ideology. It is designing so that future decisions remain open – because the market will force change anyway.
Treating security and trust as evolving systems, not one-time purchases
The fraud landscape evolves, and the ecosystem needs visibility and collaboration to keep messaging “clean and reliable”.
A sunk cost mindset treats yesterday’s controls as “good enough.” A future-oriented mindset treats trust as a continuous investment – measured, reviewed, and updated.
Optimizing for opportunity cost, not internal justification
Every euro spent to defend a failing path is a euro not spent on the capability that could actually win the next customer, enable the next integration, or protect the next revenue stream.
Closing thought
The sunk cost fallacy is not a finance mistake. It is a leadership test.
The most expensive sentence in technology is not “this is hard.” It is:
“We’ve already invested too much to stop.”
The moment you hear it – especially in messaging and telco – you should treat it like a production alarm. Not because stopping is always the answer, but because that sentence is a signal that the past is trying to run the future.
If there is one principle worth repeating at CEO level, it is this:
Sunk costs are already gone.
Strategy is what we choose to do next.
Fabrizio Salanitri
Founder & CEO, HORISEN
