The Leadership–Reality Gap: When the Presentation Improves but the Business Does Not

Leadership   /   Operational reality   /   AI

The Leadership–Reality Gap: When the Presentation Improves but the Business Does Not

The more leadership seminars and corporate discussions I attend, the more one question stays with me: Do senior leaders truly understand the operational reality of the businesses they lead?

The strategy slides are polished, the explanations sound convincing and every challenge appears to have an action plan. Yet the people dealing with suppliers, customers, systems and daily exceptions sometimes describe a very different business.

What concerns me is not presentation quality. It is the growing possibility that we are becoming better at describing improvement than at delivering it. AI makes that distinction even more important.

The view from the meeting room

This gap rarely begins with bad intentions. As people move higher in an organisation, their work naturally shifts towards reviews, performance indicators, meetings and decisions across multiple locations. At the same time, their contact with the processes that create revenue and serve customers becomes less direct.

Information is condensed as it moves upwards. A complicated issue eventually becomes a familiar set of commitments: improve efficiency, increase accountability and accelerate digitalisation.

These can be sensible objectives. But they do not tell us why a process remains slow. Is the supplier data incomplete? Does the system behave differently across plants? Is ownership unclear? Does one team have to correct another team’s input? Has anyone asked what is already being done?

Those questions may be less impressive on a slide, but they are far more useful to the business.

When existing work becomes a new strategy

In operational organisations, it is not unusual to hear an initiative announced as a new priority when teams have already spent months improving that very subject. It might concern pricing accuracy, reporting automation, faster customer responses or standardisation between systems.

The leadership objective may be right. The problem begins when existing effort is invisible and the reasons progress has stalled are not understood.

What seems to be slow execution from a distance may be a missing resource, a decision awaiting another department, unreliable master data or a technical limitation that cannot be solved by another meeting.

A serious strategy discussion should begin by asking what has already been attempted, who understands the exceptions, what actually failed and which decision would remove the constraint. Without this, the same initiative risks being relaunched each year with a new title.

AI can make weak understanding look strong

AI can structure notes, analyse reports and generate attractive executive presentations in minutes. That is valuable. It can also turn an untested assumption into language that sounds remarkably certain.

Consider a spare-parts pricing process with a high number of manual reviews. From a dashboard, it may appear to be an automation problem. An AI-generated recommendation might therefore propose reducing manual interventions.

But someone working in the process might explain that the requests are driven by missing manufacturer references, different plant structures, supplier quotations, technical alternatives, outdated costs and customer-specific agreements. Automating the visible step without addressing these dependencies could simply produce errors faster.

That is the danger: the slide becomes more persuasive while the original problem remains untouched.

“A convincing explanation is not the same as a verified explanation.”

The research reveals a gap between confidence and outcomes

This is more than a personal concern, although the research should be read carefully. These surveys measure different samples and outcomes; none, on its own, proves that senior leaders systematically misunderstand the operations they lead. Together, however, they raise important questions about trust, implementation and measurable value.

Gallup’s leadership research reported that only 20% of U.S. employees strongly agreed they trust their organisation’s leadership as of May 2026. Gallup also found that employees expressing strong leadership trust were 3.7 times as likely to be engaged. This is a U.S.-specific finding, not a global estimate. Read Gallup’s research.

McKinsey’s 2026 State of AI survey found that 80% of respondents said AI improved their individual productivity, while 37% attributed at least some positive EBIT impact to their organisation’s use of AI. These are respondents’ reported perceptions, not an audited causal assessment. McKinsey also identified fundamental workflow redesign as a distinguishing practice among the organisations capturing the strongest value. Read McKinsey’s findings.

Gartner’s October 2026 CIO survey found that only 13% reported significant or transformational value from AI. Gartner stressed the importance of defining value, knowing when to stop low-value projects and redesigning work rather than simply adding technology. Read Gartner’s announcement.

The numbers do not establish a direct cause-and-effect relationship between poor operational understanding and weak financial performance. They do, however, challenge the idea that leadership confidence, technology adoption and meeting activity should be accepted as evidence of meaningful business change.

Leadership activity is not business progress

Corporate activity is easy to count. Conferences attended, workshops organised, projects launched and steering committees completed all create the appearance of movement. Meanwhile, a customer problem can remain unresolved, an inefficient handover can stay in place, and the same data correction can be repeated month after month.

When financial targets are missed, there may be good reasons: economic uncertainty, inflation, delayed customer investment and geopolitical disruption are real influences. But if the external explanation changes every year while the same internal obstacles survive, leaders should also ask what their own decisions and operating model are contributing to the result.

A useful management review should compare the promise with the outcome. Did the response time actually improve? Were manual corrections eliminated? Did the decision become simpler? Was cost reduced without shifting it to another function? Did the customer experience change? Did margin improve sustainably?

The core business still matters

In industrial operations, commercial performance depends on details that rarely fit neatly into a boardroom graphic. A material can look interchangeable in a market-price comparison but differ technically. A new supplier price can change a transfer price. A lead-time field can be correct in one plant and outdated in another. A process can look automated in its final report while demanding hours of manual correction beforehand.

For pricing, procurement and service businesses, the consequences are substantial. Strategy can identify a margin opportunity, but only operational understanding reveals whether the product is available, equivalent, correctly costed and supported by the necessary information.

This is not an argument against senior strategy. It is an argument that strategy and operational expertise must inform each other.

Frontline challenge is not automatically resistance

One of the easiest ways to widen the gap is to interpret operational questions as unwillingness to change. Sometimes resistance is real. But sometimes the person asking the uncomfortable question is identifying the failure point before it becomes expensive.

An employee may know that a department must complete a step first, a contractual exception has been overlooked, the KPI rewards the wrong behaviour or the team lacks capacity to execute another initiative. Those observations are not reasons to abandon ambition. They are information required to make ambition achievable.

Consulting people after a decision has been announced is not the same as inviting them to test the decision before it is made.

Use AI to challenge the narrative, not decorate it

The same technology that can produce an impressive slide can also help verify it. AI could compare repeated leadership commitments with project records, inspect trends in support tickets, identify unresolved exceptions and examine whether reported efficiency gains correspond to measurable process changes.

This requires responsible access to reliable operational data, attention to privacy and permissions, and human review of the interpretations. AI outputs still need validation by people who understand the business.

A different kind of leadership conversation

Closing the gap does not mean executives must personally perform every task. It means leaders should demand a stronger relationship between decisions and execution. Before approving a major initiative, I would want the conversation to cover the following.

That last question matters. A strategy that is never tested against what happened can remain persuasive long after it has stopped being useful.

Leadership must return to the real business

Today’s leaders are expected to understand AI, regulation, geopolitical uncertainty, investor expectations and the future of work. All of these are important. None removes the need to understand how their company actually creates value.

The purpose of leadership is not to produce the most convincing explanation for performance. It is to create the conditions for better performance.

The problem is not that we have too many presentations or that technology is advancing too quickly. The problem begins when polished communication replaces curiosity about the work, and when assumptions travel further than facts.

AI can widen that distance—or help close it. The choice depends less on the quality of the slide than on the quality of the questions leaders are willing to ask.

A strategy is not successful because it looks convincing in a seminar. It is successful when the people doing the work and the customers they serve can see the difference.