Three Days Inside a Client Organisation: What the Culture Diagnostic Revealed That the Survey Never Could

I was in a client organisation for three days, commissioned to conduct a culture diagnostic that would inform their leadership team’s strategic planning for the following year. The organisation had also conducted an employee engagement survey four months earlier. The survey data was good quality: a 78% response rate, well-designed questions, sophisticated analysis including cohort breakdowns by function, level, and tenure. The CHRO had shared the results with me before I arrived, and I had read them carefully. I thought I knew broadly what I was going to find.

By the end of the first afternoon, I knew the survey had told me less than half of what I needed to understand.

This is not a critique of the survey or of the team that designed it. The survey was well-constructed and the data was honest. What it could not tell me, by its nature as a survey, was what I was beginning to observe through direct presence in the organisation: the specific texture of how people talked about the organisation in informal moments, the body language in the all-hands meeting when a particular senior leader spoke, the topics that were present in corridor conversations and absent in formal meeting rooms, the speed at which information moved across the hierarchy in different directions, and the specific interactions between two members of the leadership team that told me something important about a dynamic that had not appeared in any of the survey data. These are the things that surveys cannot capture, not because surveys are poorly designed but because they are surveys, and the most culturally significant information in an organisation is often not the information that people will write down in response to a structured question.

The survey had shown several things clearly. Manager quality scores were bimodal: a cluster of very high scores and a cluster of very low scores, with almost nothing in the middle. Psychological safety scores were high in some functions and low in others in a pattern that did not obviously correlate with business performance. Confidence in senior leadership was declining, having dropped eight points in twelve months. And tenure-based analysis showed that employees in their second and third year were less engaged than either new joiners or longer-tenured employees, a dip that is sometimes called the “sophomore slump” in engagement research and that often reflects the collision between initial enthusiasm and the reality of the organisation’s actual culture.

All of this was useful. But it was descriptive rather than explanatory. It told me what was happening. It did not tell me why. And the “why” is always the strategically important question, because the interventions that address the symptom without understanding the cause produce temporary improvements that revert within twelve months. THE DIAGNOSTIC WORK THAT PRODUCES DURABLE RECOMMENDATIONS requires understanding not just the shape of the problem but the mechanism that is generating it, and the mechanism is almost always more complex and more contextually specific than the survey data alone can reveal.

On the first morning, I attended the weekly operations review. I said nothing; I simply observed. The meeting was formally structured, well-facilitated by the COO, and operationally focused. What I observed was this: three of the eight people in the room spoke freely and at length. Two spoke when directly addressed. Three said almost nothing. The three who said nothing were not passive personalities; I had read their profiles, and their backgrounds described people with strong views and significant experience. The pattern of who spoke and who did not was not random, and it did not correspond to functional relevance: the topics discussed were equally relevant to everyone in the room.

By the end of the first day, I had identified three specific dynamics that the survey data had not surfaced and that were, in my assessment, more explanatory of the engagement patterns than any of the variables the survey had measured. I will describe them in the order in which they became visible.

The first was a specific power dynamic between two members of the senior leadership team, which was structuring the behaviour of the broader team in ways that were highly predictable once you knew the dynamic existed. One leader, I will call her the functional head, was substantively more influential in informal conversations and corridor exchanges than her formal authority would suggest. A second leader, I will call him the operations head, was formally more senior but less influential in the informal network. The three people who said nothing in the operations review were all in the functional head’s informal network. They did not speak in formal forums because they had learned that their views found their way to decision-making through a different channel. THE INFORMAL POWER STRUCTURE WAS NOT PATHOLOGICAL; it was, in a sense, functional. But it was creating a visible split between formal and informal authority that was generating confusion and disengagement in the layers below.

The second observation came from three informal conversations with employees in their second and third year, the “sophomore slump” cohort that the survey had flagged. In each conversation, independently and without prompting, the same theme emerged: they had joined the organisation because of what they understood its culture to be from the recruitment process and their first weeks on the job, and they had spent the following eighteen months discovering that the actual culture in operational reality was different from the projected culture in recruitment and onboarding. Not dishonestly different; the values the organisation projected were genuine values, held by most of the senior leaders. But THE CULTURE AS EXPERIENCED AT THE MANAGER LEVEL was shaped by a different set of pressures: delivery pressure, margin pressure, the implicit norm of moving fast and not asking too many questions about the how. The values and the operational culture were in tension, and the second-year employees were the cohort most acutely aware of it because they were far enough in to have lost the protective optimism of new joiners but not yet far enough in to have normalised the gap.

The third observation was structural and required the full three days to see clearly. I tracked, informally but deliberately, how quickly different categories of information moved across the organisation, and in which directions. Positive information, new client wins, product milestones, operational achievements, moved upward quickly and laterally quickly. It was the currency of the formal communication channels. Problem information, delivery challenges, quality issues, concerns about strategic direction, moved upward slowly, selectively, and in some cases not at all. The formal communication channels were rich with positive information and sparse with problem information in a ratio that was not consistent with an organisation facing the level of operational complexity this organisation was dealing with.

This pattern is not unusual; it is one of the most common culture signatures of an organisation with insufficient psychological safety at the middle management level. But its significance in this organisation was specific: the senior leadership team, which was genuinely committed to building an open and candid culture, was making strategic decisions based on a positive information bias that they were not aware of, because the filter through which information reached them was invisible to them. THE ENGAGEMENT SURVEY HAD NOT SURFACED THIS because the psychological safety questions had been scored against the abstract idea of safety, not against the specific experience of raising a problem up the hierarchy, which is the dimension that matters most for strategic decision quality.

The combination of these three observations, the informal power structure, the values-operations gap at the manager level, and the information asymmetry between positive and negative signals, gave the culture diagnostic a specificity and an explanatory power that the survey data alone could not have provided. The recommendations that emerged from the diagnostic were different from the recommendations that a survey-only process would have generated, more specific, more contextually grounded, and more directly linked to the mechanisms that were generating the culture patterns the organisation was concerned about.

The recommendations I brought to the leadership team at the end of the three days addressed the mechanisms rather than the symptoms. The bimodal manager quality scores were a symptom of a leadership selection and development process that had prioritised technical expertise over people leadership capability for the previous three years of rapid growth. The recommendation was not a management training programme but a redesign of the leadership selection criteria and a structured management capability audit that would enable targeted intervention where it was most needed.

The values-operations gap was a symptom of a performance management system that rewarded delivery speed without calibrating for how delivery was achieved. The recommendation was a specific change to the performance conversation structure for managers, combined with explicit senior leader modelling of the values in situations where they were in tension with delivery pressure, made visible to the broader organisation through deliberate communication. THE SECOND-YEAR ENGAGEMENT DIP would be addressed at source rather than at symptom, not through onboarding enhancement (the usual intervention) but through a fundamental recalibration of the gap between what the organisation projected and what it delivered at the manager level.

The information asymmetry recommendation was the most structural of the three: a redesign of the executive team’s information diet, specifically the introduction of regular direct sensing mechanisms that bypassed the filter, combined with explicit signals from senior leaders that problem information was not only safe but valued, made concrete through visible positive responses to the first examples of upward candour that the new signals would generate.