The 70% Myth: Why CEO Standards, Not Manager Talent, Drive Employee Trust

Two directors report to the same vice president. Both lead teams of twelve. One runs a disciplined operation: expectations are documented, commitments are followed through, missed deadlines are addressed, and conduct problems do not drift for months without consequence.

The other director runs on instinct. Favorites get flexibility. Problems are handled “case by case.” Expectations change depending on pressure, personality, or timing. Employees are never quite sure what will happen if they raise a concern, miss a deadline, or challenge a decision.

By the time engagement scores come back, the difference between the two teams is unmistakable. One team reports higher trust, clearer priorities, and stronger confidence in leadership. The other reports confusion, inconsistency, and declining belief that concerns will be handled fairly.

The leadership team looks at the data and reaches the familiar conclusion: one director is a strong manager, and one is not.

HR is asked to fix the weaker manager. Coaching is discussed. A development plan is drafted. Perhaps the director is sent to leadership training. The organization treats the gap as a manager-capability problem.

What almost nobody asks is why the company allowed two directors, reporting to the same executive, to operate under two different management standards in the first place.

That is the question the CEO owns.

The Number Everyone Cites

Gallup’s widely repeated finding is clear: managers account for at least 70% of the variance in employee engagement scores across business units. The statistic has become one of the most quoted claims in modern management.

The usual conclusion is equally familiar. If managers explain most of the variation in engagement, then organizations need to select better managers, train existing managers, and develop stronger people leaders.

That conclusion is not wrong. Manager quality matters. A weak manager can damage trust, reduce productivity, increase turnover risk, and make even a strong strategy feel chaotic at the employee level.

But the conclusion is incomplete.

The 70% statistic tells leaders where engagement variance shows up. It does not fully explain why manager behavior varies so widely inside the same organization.

That distinction matters. If the CEO interprets the number as a talent problem, the organization will invest primarily in manager selection and development. If the CEO interprets the number as a standards problem, the organization will ask a harder question: why are managers being allowed to define, bend, or ignore the rules on their own?

What the Statistic Does Not Tell Us

The standard interpretation of Gallup’s number places the manager at the center of the engagement problem. That is understandable. Employees experience the company most directly through their manager. The manager assigns work, clarifies expectations, responds to conflict, recognizes contribution, corrects behavior, and decides whether issues are addressed or avoided.

But a manager is not operating in a vacuum.

The statistic does not tell us whether every manager was working under the same enforceable expectations. It does not tell us whether the organization corrected managers who ignored those expectations. It does not tell us whether HR had the authority to intervene when a manager’s behavior drifted from policy. It does not tell us whether the variance was caused by managerial talent, or by the amount of discretion managers were given to invent their own rules.

Those are not small omissions. They are central to the problem.

A company can have excellent policies and still tolerate inconsistent management. It can have leadership values and still allow different teams to experience different consequences for the same behavior. It can have HR processes and still leave HR with responsibility for cleanup rather than authority over prevention.

In that environment, the manager becomes the primary variable because the organization has failed to define anything stronger above the manager.

Manager Variance Is Often CEO Standard Variance

When a CEO does not define what good management means in enforceable terms, every manager fills the gap with personal judgment.

Some judgment is excellent. Some is inconsistent. Some is avoidant. Some is self-protective. Some is shaped by favoritism, fear of conflict, pressure to hit numbers, or reluctance to challenge a high performer.

The result is variance. Not because the organization has discovered a mysterious shortage of management talent, but because it has created too much room for individual interpretation.

This is the CEO standard gap.

It appears when one manager documents performance issues while another lets the same behavior continue informally. It appears when one employee is disciplined for conduct that another employee is allowed to repeat. It appears when one complaint is escalated promptly while another is softened, delayed, or handled quietly because the accused manager is valuable to the business.

Employees see these differences. They do not need a survey to detect them.

They notice when the rule depends on the manager. They notice when consequences depend on status. They notice when HR says one thing, but executives permit something else. Over time, they stop judging the organization by its stated values and start judging it by its tolerated exceptions.

That is when trust begins to decline.

The Mechanism Behind the Trust Problem

Employee trust does not erode only because a manager lacks skill. It erodes when employees conclude that the organization cannot be counted on to apply its own standards.

The mechanism is straightforward:

The CEO does not define an enforceable management floor.

Managers improvise.

HR is asked to support, coach, and document, but not always empowered to stop the drift.

Employees experience inconsistency across teams.

Trust declines because fairness appears conditional.

Engagement scores then report the variation as a manager problem.

By the time the data reaches the executive team, the original failure has been disguised. The survey points to the manager. The development budget follows the manager. The CEO remains one level removed from the problem, even though the conditions that produced the variance were created at the top.

That is why manager training alone so often disappoints.

Training can improve skill. It can give managers better language, better tools, and better confidence. But training cannot replace an enforceable standard. It cannot make consequences consistent if executives are unwilling to apply them. It cannot give HR authority if the CEO has not granted it. It cannot stop a high-performing manager from violating expectations if the organization is still willing to protect results over conduct.

Training raises capability. Standards reduce variance.

The distinction is critical.

Why the Technical Fix Falls Short

Most organizations respond to manager variance with technical solutions. They introduce leadership development programs. They revise manager competencies. They add coaching. They improve engagement surveys. They build learning paths for difficult conversations, feedback, delegation, and performance management.

These efforts have value. Managers need skill. Many have never been properly prepared for the role. Some were promoted because they were strong individual contributors, not because they knew how to lead people.

But the deepest problem is not always skill. It is discretion without accountability.

A manager can be trained and still avoid difficult conversations. A manager can know the policy and still apply it selectively. A manager can attend a leadership program and still protect a favorite employee. A manager can understand documentation and still delay using it because doing so would create conflict.

The question is not only whether managers know what to do.

The question is whether the organization requires them to do it.

That requirement cannot live only in training materials. It has to appear in operating expectations, escalation rules, performance reviews, promotion decisions, compensation conversations, and executive follow-through.

A standard only becomes real when violating it has a consequence.

What a Real Management Standard Requires

If CEOs want less variance in employee trust, they need to define the management floor beneath every leader. That floor is not a slogan. It is a set of enforceable expectations for how people are managed.

At minimum, the CEO standard should answer five questions:

• What must every manager clarify before holding employees accountable?

• What must every manager document when performance or conduct begins to drift?

• What issues must be escalated, and by when?

• What discretion does a manager have, and where does that discretion end?

• What consequence applies when a manager ignores the standard?

Those questions change the conversation. They move the organization away from vague expectations about “good leadership” and toward an operating model for managerial conduct.

The standard should apply to the well-liked manager and the difficult manager. It should apply to the manager who delivers results and the manager who struggles. It should apply when the issue is inconvenient, politically sensitive, or attached to someone with influence.

Otherwise, employees will learn the real rule: standards exist until enforcing them becomes uncomfortable.

The Role HR Can and Cannot Play

HR can build the architecture. HR can define the policy, create the documentation process, train managers, monitor patterns, and warn executives when practices become inconsistent.

But HR cannot substitute for CEO authority.

This is where many organizations quietly fail. They assign HR responsibility for engagement, culture, employee relations, and manager capability without giving HR the standing to close the gaps that create distrust. HR is expected to support managers, advise leaders, respond to complaints, and reduce risk. But when a manager with strong business results repeatedly violates the standard, HR often has to persuade rather than require.

That is not a governance model. It is a hope strategy.

If HR can identify the risk but cannot compel correction, the organization has not created accountability. It has created documentation of avoidable failure.

The CEO has to decide whether HR is merely a support function or whether it has authority to flag, escalate, and require action when management practices violate the company’s standards.

Without that authority, HR becomes the department asked to repair trust after leadership has already spent it.

A Case Pattern CEOs Should Recognize

Consider a mid-sized organization with strong growth, uneven manager capability, and rising employee complaints. The executive team believes the issue is a handful of difficult managers. HR believes the issue is inconsistency. Employees experience it as unfairness.

On one team, attendance problems are documented quickly. On another, they are ignored until resentment builds. One manager escalates conduct concerns immediately. Another handles them informally for months. One department uses performance expectations carefully. Another relies on vague warnings and personality-based judgments.

The CEO initially sees multiple people problems. The better diagnosis is one standards problem.

The practical intervention is not to send every manager through another generic leadership program. The intervention is to define the management floor:

Every manager must set expectations in writing when performance begins to drift.

Every repeated conduct issue must be escalated within a defined timeframe.

Every exception must be documented and reviewed.

Every manager is evaluated not only on results, but on whether those results were achieved within the company’s management standards.

HR has standing to flag noncompliance directly to senior leadership.

That kind of operating discipline does not eliminate the need for manager training. It makes training matter. Managers are no longer learning techniques in isolation. They are being trained into a standard the organization is prepared to enforce.

That is where trust begins to change.

What Employees Are Really Responding To

Employees do not experience “manager variance” as a statistical concept. They experience it as daily inconsistency.

They notice whose behavior gets corrected and whose behavior gets excused. They notice whether complaints are handled differently depending on who is involved. They notice whether a manager’s title, tenure, revenue, or relationship with senior leaders changes the rules.

Once employees believe standards depend on the manager enforcing them, trust becomes harder to rebuild. The issue is no longer one bad supervisor or one missed conversation. The issue becomes organizational credibility.

This is why the familiar phrase “people leave managers, not companies” is only partly true.

People may leave because of a manager. But they also leave because the company allowed that manager to operate without correction. They leave because the organization saw the pattern and did not stop it. They leave because HR knew, employees knew, and leadership still treated the issue as a coaching matter rather than a standards failure.

The manager may be the point of impact.

The company is still the system of permission.

The CEO Agenda

If CEOs want to reduce engagement variance, they should not abandon manager development. They should place it inside a stronger operating system.

The CEO agenda is clear:

Define the management floor. Make explicit what every manager must do when setting expectations, addressing performance, documenting concerns, escalating complaints, and applying consequences.

Reduce unauthorized discretion. Managers need judgment, but not unlimited freedom to apply standards differently across teams.

Give HR standing. HR must be able to flag management drift before it becomes a complaint, claim, resignation, or trust failure.

Make exceptions visible. If a leader approves an exception, it should be documented, reviewed, and understood as an executive decision, not hidden inside managerial discretion.

Tie management conduct to advancement. No manager should be promoted solely because they deliver numbers while leaving trust, documentation, or conduct problems behind them.

Enforce the standard when it is costly. The true test is not whether the CEO supports standards in principle. It is whether the CEO enforces them against a manager who is otherwise valuable.

That last point is where the culture becomes real.

The 70% Statistic Is Real. The Usual Interpretation Is Too Small.

Gallup’s 70% figure should not be dismissed. Managers matter enormously. They shape the daily employee experience more directly than almost anyone else in the organization.

But the statistic should not be used to let CEOs off the hook.

If managers account for most of the variance in engagement, CEOs should ask why the organization has allowed so much variance in management practice. Why are employees having radically different experiences under the same company values? Why do consequences change by department? Why does escalation depend on the confidence or comfort level of an individual manager? Why is HR expected to fix problems it was not authorized to prevent?

The answer is rarely found in manager talent alone.

It is found in the standards the CEO defined, the exceptions the CEO allowed, the authority the CEO gave HR, and the consequences the CEO was willing to enforce.

Manager talent matters.

But CEO standards set the floor.

When that floor is clear, enforced, and applied without exception, manager variance narrows. Employees may still prefer one manager over another, but they are less likely to experience the organization as arbitrary. They know the rule does not depend entirely on who supervises them.

When the floor is weak or undefined, every manager becomes a separate culture. Every team becomes its own interpretation of fairness. Every employee learns to ask not “What is the company standard?” but “Who is managing this?”

That is not a manager problem alone.

It is a CEO standard gap.

The 70% statistic does not prove that managers determine employee trust by themselves. It proves that too many organizations have left employee trust almost entirely in their hands.