The Conversation the CHRO Needs to Have with the CEO but Cannot: Notes on the Pattern

I was three months into a CHRO advisory engagement when the CHRO I was working with said something that I have heard, in various formulations, from almost every CHRO I have advised over the past decade. She said: “I know what needs to be said. I know to whom it needs to be said. I know when it needs to be said. And I cannot figure out how to say it.” The conversation she was describing was with her CEO. The subject was the CEO’s own leadership behaviour, specifically a pattern of decision-making that was undermining the executive team’s confidence and creating a growing talent risk in the layer directly below the C-suite. She had the data. She had observed the pattern across multiple contexts over nine months. She had evidence from the engagement survey, from exit interview themes, and from direct conversations with three members of the executive team who had approached her separately.

What she did not have was a way to have the conversation. Not the courage; she had that. Not the relationship; she had that too, built over two years of genuine mutual respect and demonstrated professional capability. What she did not have was a clear answer to the question that was stopping her: how do you tell a CEO something about their own leadership that they have not asked to hear, in a way that is honest, respectful, preserves the relationship, and actually has a chance of changing something?

I want to write about this pattern specifically, because it comes up in almost every CHRO advisory engagement I have been part of, and because the advice available to CHROs on this question is, in my view, consistently insufficient. It focuses on courage, on preparation, on framing. These are all relevant. But they do not address the structural and relational complexity that makes this specific conversation uniquely difficult, and they do not give the CHRO the practical thinking architecture they need to design a conversation that has a genuine chance of producing a useful outcome.

The difficulty of this conversation is structural before it is relational or psychological. The CHRO is in an unusual position relative to the CEO: they are both a subordinate (they report to the CEO, their role exists within the CEO’s authority) and a steward of the organisation’s health in ways that sometimes require them to hold the CEO to account. They are simultaneously someone who serves the CEO and someone who is responsible, in a meaningful institutional sense, for the quality of leadership at the top of the organisation, which includes the CEO’s own leadership. That dual position does not have a clean resolution; it is a structural tension that the CHRO must navigate every day.

The conversation about the CEO’s behaviour adds another layer of complexity. The CHRO is bringing data that the CEO has not requested, about behaviour that the CEO may not be aware of, with an implicit or explicit invitation to change. In most professional contexts, the person with the authority to hold another person to account and invite them to change is the person with greater positional authority. Here, the positional authority runs the other way. THE CHRO IS ASKING THE CEO TO HEAR SOMETHING THAT THE CEO IS UNDER NO INSTITUTIONAL OBLIGATION TO HEAR, and is doing so from a position of lower positional authority, which means the CHRO is entirely dependent on the quality of the relationship and the credibility they have built to create the conditions under which the CEO is willing to engage with the data at all.

The third layer of structural complexity is timing. The CHRO who has been accumulating data about a CEO behaviour pattern over months has typically been in a gradually worsening position: the longer they wait, the more data they have, but also the more the situation is deteriorating, and the more the CHRO’s own credibility is at risk for having not surfaced it earlier. There is no clean moment to have this conversation, which means the CHRO ends up waiting for a moment that feels right while the right moment does not arrive, and the situation continues to evolve.

In the advisory conversations I have had with CHROs who are carrying this specific challenge, I have observed a consistent set of patterns in the period before the conversation occurs. Understanding these patterns is useful because they describe the conditions that make the eventual conversation more or less likely to go well.

The first pattern is progressive data accumulation without progressive framing. The CHRO is doing the observational work, gathering evidence from multiple sources, tracking the pattern carefully. But they are not doing the parallel work of developing a clear and honest frame for the conversation, which means the gap between the data they are accumulating and their ability to use it productively is growing rather than closing. The more data they have, the harder it becomes to decide what to include and what to leave out, what to foreground and what to contextualise. A CLEAR FRAME DOES NOT EMERGE AUTOMATICALLY FROM ACCUMULATING DATA; it requires a separate act of deliberate thinking that most CHROs have not had the structural support to do.

The second pattern is relational protection at the expense of organisational responsibility. The CHRO has a relationship with the CEO that they value and that enables the function’s effectiveness. They are reluctant to put that relationship at risk by saying something that the CEO might receive defensively. This is an understandable priority, but it is one that gradually compromises the CHRO’s core institutional function. The CHRO who does not surface material leadership concerns because they are protecting the relationship is prioritising the instrument over the purpose: the relationship is valuable because it enables the CHRO to do the job, and the job includes surfacing material concerns about leadership effectiveness when the evidence requires it.

The third pattern, the one I find most important, is the conflation of the conversation with a judgment. The CHRO is treating the data-sharing conversation as if it were a verdict or a performance assessment: I am going to tell the CEO that they are failing at X. The conversation is then charged with a weight it does not need to carry. THE MORE PRODUCTIVE FRAME IS EXPLORATORY: I have observed a pattern, I have data that I think is important for you to have, and I want to understand whether you are seeing what I am seeing. That frame is both more accurate (the CHRO does not have the full picture, and the CEO’s perspective is genuinely valuable) and more likely to produce a receptive rather than defensive response.

Back to the CHRO I was working with. Over the course of three advisory sessions, we worked through the design of the conversation together, and I want to share what that design process involved because it is the element that is most consistently absent from the advice that CHROs receive on this subject.

The first step was to agree on the purpose of the conversation with precision. Not: to tell the CEO what is wrong. Not: to hold the CEO accountable. The precise purpose was: to share data that the CEO needs in order to make informed decisions about their own leadership in the context of the organisation’s current strategic priorities, and to invite the CEO’s perspective. That purpose is honest, it is respectful, and it is achievable in a single conversation. It does not require the CHRO to be right about everything, and it does not require the CEO to agree with everything. It creates the conditions for a productive exchange rather than a confrontation.

The second step was to select the data carefully. Not all of the data the CHRO had accumulated was necessary or useful in this conversation. The selection principle was: what is the smallest amount of evidence that is most directly relevant to the impact the CHRO is concerned about, expressed in the language and metrics that the CEO cares about? The aim was three specific observations, each grounded in data the CEO could verify, rather than a comprehensive account of everything the CHRO had observed over nine months. SELECTIVITY IS NOT DISHONESTY in this context; it is the design discipline that makes the conversation receivable rather than overwhelming.

The third step was to prepare for the CEO’s range of responses, not just the defensive response. The CHRO had been preparing primarily for the scenario where the CEO pushed back, denied the observation, or responded with frustration. We also prepared for the scenario where the CEO was genuinely unaware of the pattern and was affected by hearing it, for the scenario where the CEO already knew and had been thinking about it, and for the scenario where the CEO asked what the CHRO thought should happen next. Having a prepared response to all of these scenarios changed the CHRO’s emotional state going into the conversation: they felt prepared rather than exposed, which changed the quality of their presence in the room.

The CHRO I described had the conversation in the fourth month of our advisory engagement. It lasted ninety minutes. It went well. Not dramatically: the CEO did not have a transformative moment of self-recognition in the room. But the CEO listened, asked good questions, provided context that the CHRO had not had, and at the end of the conversation, committed to three specific changes in their approach to a specific category of decisions. Six months later, the CHRO reported that two of the three changes were visible in the CEO’s behaviour and were being noticed by the executive team. The talent risk, which had not yet become a talent loss, was stabilising.

I share this not as a success story, though the outcome was positive, but as an illustration of what becomes possible when the conversation is designed rather than simply summoned from courage. The CHRO’s courage was never in question. What changed was the quality of the preparation: the clarity of purpose, the discipline of data selection, and the readiness for the full range of responses. Those elements can be built through a structured advisory relationship. THEY DO NOT EMERGE RELIABLY FROM SELF-PREPARATION ALONE, because the structural isolation the CHRO is in makes it genuinely difficult to think clearly about a situation in which they are so personally and professionally invested.