Marketing has become extraordinarily good at making uncertainty visible.
Dashboards update continuously. Every impression, click, lead and sale is counted. Campaign performance can be divided by audience, channel, creative, geography, hour and device. When a number moves, there is usually a person expected to explain why.
This creates the appearance of control.
If performance improves after a campaign change, the intervention is treated as the cause. If performance deteriorates, the person managing the account begins searching for an error. Perhaps the bid was changed too quickly. Perhaps the audience was too broad. Perhaps the creative should have been replaced sooner.
Some of this scrutiny is necessary. Marketing decisions should be examined, and practitioners should be accountable for the quality of their work.
The problem begins when accountability is confused with causality.
A marketer may be responsible for managing an outcome without controlling all the forces that produce it. Paid media specialists, in particular, operate inside systems shaped by consumer demand, competitors, platform algorithms, auction dynamics, pricing, stock availability, economic conditions and chance.
Yet many organisations continue to evaluate them as though the relationship between action and outcome were direct.
This is not merely a measurement error. It is a management problem with potentially serious psychological consequences.
Measurement has expanded faster than control
The growth of marketing technology has produced an important but often overlooked asymmetry.
Organisations can observe far more than they can control.
A paid media manager can see that conversion rates declined at 2pm on Tuesday. That does not mean the manager can identify why it happened, still less that they caused it.
The decline might reflect a change in competitor pricing, a shift in consumer demand, a technical issue on the website, a delayed conversion, a product becoming unavailable or a change within the advertising platform itself.
Even Google's guidance on performance fluctuations acknowledges that results can be affected by market dynamics, competition, seasonality, conversion delays and account changes.
The dashboard presents all these effects as a single number.
Someone must then turn that number into a story.
This creates pressure to infer causality from sequence. A campaign change occurred, followed by a movement in performance, so the change becomes the explanation. If no obvious intervention preceded the movement, the person managing the account is expected to find one.
The result is often false precision. Uncertainty is translated into a confident narrative because "we do not yet know" is considered an inadequate response.
More data can therefore create an illusion of understanding. It allows an organisation to describe a change in extraordinary detail without necessarily explaining its cause.
Why people overestimate their influence
In 1975, psychologist Ellen Langer introduced the concept in her paper "The Illusion of Control".
Langer examined how people can behave as though they have influence over outcomes that are partly or entirely determined by chance. The perception of control becomes stronger when a situation contains familiar characteristics of skill, such as choice, competition, active involvement and specialist knowledge.
Modern marketing contains all of these characteristics.
Marketers choose audiences, allocate budgets, write adverts, set bids and design experiments. They compete with other advertisers and receive immediate feedback on their decisions.
Their actions genuinely matter. But the presence of skill does not eliminate uncertainty.
This is what makes the illusion so persuasive. Paid marketing is neither entirely controllable nor entirely random. It combines judgement with probability. The practitioner can improve the likelihood of a favourable result without being able to guarantee one.
The distinction is easy to understand intellectually and difficult to maintain emotionally.
When a campaign performs well, the marketer is encouraged to view the outcome as evidence of competence. When it performs badly, the same logic turns against them. If success demonstrates skill, failure must demonstrate its absence.
Over time, commercial performance can become entangled with personal identity.
Self-blame offers the comfort of an explanation
Self-blame is painful, but it can also be psychologically reassuring.
If a negative outcome was caused by a personal mistake, then the world remains understandable. The mistake can be identified, corrected and avoided next time.
The alternative is more unsettling.
The marketer may have made a reasonable decision and still received a poor result. Several forces may have changed simultaneously. The available data may not support a clear conclusion. It may simply be too early to distinguish a lasting pattern from ordinary variation.
Accepting this means accepting the limits of one's knowledge and influence.
Self-blame replaces that uncertainty with a coherent account:
I caused the problem, therefore I can solve it.
This belief encourages action, even when further intervention is unlikely to help. Campaigns are restructured, budgets are moved and targeting is changed before enough evidence has accumulated.
Each intervention makes the system more difficult to interpret. The next performance movement becomes harder to attribute, which creates more uncertainty and invites further intervention.
What appears to be energetic optimisation may, in reality, be an attempt to manage anxiety.
The practitioner is no longer changing the campaign because the evidence supports a change. They are changing it because action temporarily restores a sense of control.
Agencies institutionalise the problem
The agency model can turn this individual bias into an organisational one.
Agencies are frequently appointed on the basis of outcome-oriented promises: increased revenue, lower acquisition costs, improved returns or accelerated growth.
Those promises are commercially attractive because they sound decisive. They also conceal the number of conditions required to achieve them.
An agency may manage media investment without controlling the client's pricing, product quality, stock levels, website, customer service, sales process or promotional strategy. It may be unable to see changes being made elsewhere in the business.
Nevertheless, when revenue declines, responsibility tends to move towards the most visible performance channel.
This produces a structural mismatch between accountability and authority.
The agency is expected to answer for an outcome produced by a system it only partly controls. That expectation is then transferred to the practitioner closest to the account.
The weekly reporting cycle intensifies the pressure. Short-term movements must be explained before there is enough evidence to interpret them. The client expects an answer. The account director needs a narrative. The specialist is encouraged to convert uncertainty into certainty.
A decline caused by several interacting factors becomes "creative fatigue". A temporary variation becomes an "audience issue". A wider decline in demand becomes a campaign management failure.
These explanations are not necessarily dishonest. They may be the best available hypotheses. The problem is that hypotheses are often communicated as facts because the agency has sold confidence rather than judgement.
The industry then rewards certainty theatre. The person who offers a simple explanation appears more competent than the person who accurately describes the limits of the evidence.
The psychological cost of responsibility without control
The distinction between responsibility and control matters because perceived control has a significant relationship with how people experience stress.
Research on occupational health has repeatedly examined the combination of high demands and low decision-making control, often described as job strain.
A systematic review and meta-analysis of psychosocial occupational exposures found that high job strain was associated with a greater risk of depression. Lower job control was also independently associated with increased risk.
A separate systematic review of workplace stressors found associations between diagnosed mental-health-related sickness absence and factors including high psychological demands, low job control, low reward and effort-reward imbalance.
These findings should not be interpreted as meaning that a difficult advertising account will cause a mental health condition. The point is broader.
Work becomes psychologically hazardous when demands are persistently high, control is limited and the individual is still expected to accept personal responsibility for the outcome.
Paid media can reproduce precisely these conditions.
Performance is visible at all times. The commercial pressure is immediate. The practitioner is encouraged to monitor continuously, yet many of the forces shaping the result remain outside their influence.
This can produce a state of permanent vigilance.
The dashboard is checked before the working day begins and after it ends. A poor result becomes difficult to leave at work because the account continues to operate overnight. The next client conversation is mentally rehearsed before the evidence is complete.
The practitioner may begin to experience each movement in performance as a judgement of professional competence.
The World Health Organization defines burnout as an occupational phenomenon resulting from chronic workplace stress that has not been successfully managed. It identifies exhaustion, increased distance or cynicism towards work, and reduced professional efficacy as its three dimensions.
The final dimension is particularly important.
When people are repeatedly held responsible for outcomes they cannot reliably control, they may not conclude that the system is unpredictable. They may conclude that they are ineffective.
The movement from control to helplessness
The illusion of control can eventually produce its opposite.
Initially, the practitioner assumes that every poor outcome can be corrected through greater effort. They monitor more frequently, work longer hours and make more interventions.
But if effort and outcome remain weakly connected, their sense of agency may begin to erode.
Psychological research has long examined the effects of uncontrollable stress. Although the scientific understanding of learned helplessness has developed considerably since its earliest formulations, perceived controllability remains an important component of stress and coping research.
A review of the characteristics of psychological stress notes that whether a stressor is perceived as controllable can influence behavioural and stress responses.
This creates the possibility of moving between two equally distorted positions.
The first is excessive control:
Every result is a reflection of my decisions.
The second is excessive helplessness:
Nothing I do makes any difference.
Neither provides a useful basis for professional judgement.
Marketing decisions matter. Better creative, stronger measurement and more disciplined experimentation can improve the probability of success.
But probability is not certainty, and influence is not control.
A mature marketing function must be able to hold both ideas at once.
A more accurate model of accountability
Organisations need a more precise definition of what marketers should be accountable for.
The practitioner should be accountable for the quality of the decision, not for guaranteeing the outcome.
This includes the quality of the evidence used, the clarity of the hypothesis, the appropriateness of the intervention and the discipline with which the result is evaluated.
Outcomes still matter. Commercial functions cannot excuse consistently poor performance by appealing to uncertainty.
But a single outcome cannot reveal whether a decision was good.
A well-designed experiment can fail. A badly designed campaign can temporarily succeed. Only repeated observation allows the organisation to distinguish skill from luck and process quality from short-term variation.
This requires separating three forms of control.
Direct control
These are decisions the marketing team can make itself: budgets, campaign structures, bids, targeting, creative, measurement design and testing priorities.
Influence
These are outcomes marketing can affect but not determine: attention, consideration, conversion probability, customer acquisition and revenue.
Exposure
These are conditions the team must monitor but cannot directly manage: competitor behaviour, platform changes, economic conditions, seasonality, changes in customer demand and events elsewhere in the business.
The purpose of this distinction is not to reduce accountability. It is to locate it correctly.
Marketers should be challenged on the quality of decisions within their control and on how intelligently they respond to conditions outside it.
They should not be required to pretend that the final commercial outcome is entirely theirs to command.
From performance certainty to decision quality
A more psychologically sustainable marketing culture would also produce better commercial decisions.
Forecasts would be expressed as ranges rather than promises. Assumptions would be made explicit. Teams would distinguish observed changes from causal explanations.
Reporting conversations would begin with better questions:
What changed?
How confident are we that the change is meaningful?
Which possible causes are supported by evidence?
Which factors were within our control?
What action is justified now?
What evidence should we wait for before intervening?
Agencies would explain the conditions on which their forecasts depend. Clients would understand that hiring expertise improves the quality of decisions, not the predictability of the market.
Leaders would resist the temptation to reward the person with the most confident explanation. They would create room for uncertainty without allowing uncertainty to become an excuse for weak thinking.
This is a more demanding standard than conventional accountability.
It requires practitioners to make their reasoning visible. It requires managers to evaluate decisions over an appropriate period. It requires agencies to stop promising a degree of control that the operating system cannot provide.
Most importantly, it separates professional responsibility from personal blame.
Expertise is knowing the limits of control
Good marketers care deeply about results. They should.
But seriousness about performance does not require the belief that every result is personally controllable.
A practitioner can make an intelligent decision that produces a disappointing outcome. They can miss an important signal without being incompetent. They can say that the evidence is inconclusive without avoiding responsibility.
The function of expertise is not to eliminate uncertainty.
It is to make better decisions within it.
Marketing technology has given organisations unprecedented visibility. It has not given them command over customer behaviour, competition or the wider economy.
The illusion of control emerges when we confuse the ability to observe a system with the ability to govern it.
That confusion damages decision-making. It also places an unreasonable psychological burden on the people expected to explain every movement in performance.
The solution is not lower standards or weaker accountability.
It is a more intelligent form of accountability, one that recognises the difference between decisions, influence and outcomes.
The best marketers do not control the market. They understand where their influence begins, where it ends and how to act responsibly in the space between.