This is an anonymised account. The organisation is not named, and details that could identify it have been left out or generalised. The engagement ended after the diagnostic phase: what follows is what the diagnostic found, not the results of an intervention.
The organisation's view of itself
The organisation is an established Indian engineering manufacturer operating across several sites. Its senior leaders saw the culture as open: a place where people could raise problems, challenge decisions and bring ideas forward. Its stated values said the same.
They came to us with problems they did not connect to that culture:
- Middle managers were leaving faster than the organisation could replace their experience, and the reasons recorded on exit were generic.
- Engineering and operations worked in silos, and improvements made in one part of the business were slow to reach the rest.
- Earlier leadership programmes had not changed behaviour, despite real investment in them.
The working assumption was that managers needed better skills.
What we measured
Before recommending anything, we ran an organisational diagnostic using our EQ Pulse framework, covering leadership climate, communication, psychological safety and readiness for change across the sites. Alongside it, a cohort of senior and middle managers completed EQ Signal behavioural profiles, giving each of them an individual baseline.
What the diagnostic found
1. Escalation was the safe choice
Decisions piled up with senior site leaders. At first this looked like leaders holding on to control. It was not. Managers below them had learned that taking a decision on their own carried a reputational risk, while escalating it carried none. People were not being told to stay quiet; they had worked out that staying quiet, and passing decisions up, was what kept them safe.
That is the opposite of the culture the organisation believed it had.
2. The stated values and the behaviour did not match
The diagnostic surfaced a consistent gap between the organisation's stated values and how leaders actually behaved, sharpest at the level just below the top. Leaders said they welcomed challenge. The people reporting to them did not experience challenge as welcome.
3. Strategy was lost in translation, not in design
Senior leaders shared a coherent strategy. The breakdown happened in the middle layer, where that strategy was turned into day-to-day priorities differently at each site. In a culture where checking upward was safer than interpreting, each site filled the gap its own way.
4. People were not leaving for money
The reasons recorded on exit were generic. The diagnostic told a more specific story: capable managers felt unable to grow as leaders in a culture that rewarded execution and penalised initiative. They were leaving to find room to lead.
5. Collaboration broke down at the boundaries
The managers' profiles showed empathy and conflict-resolution scores noticeably lower in cross-functional situations than within their own teams. Engineering and operations also ran on different meeting rhythms, reporting lines and success measures. Asking them to collaborate without changing that structure was asking for goodwill to overcome the system.
The organisation saw a skills problem. The data showed a safety problem: people had learned that the safest thing to do was to say little and decide nothing.
What this means for AI adoption
We did not measure AI readiness in this engagement. But the conditions the diagnostic found are exactly the ones that make AI adoption fail quietly.
Adopting AI well depends on people who will try a new tool before they are sure of it, say when its output looks wrong, admit when they got something wrong while learning, and make judgement calls rather than wait to be told. In an organisation where initiative carries a reputational risk and escalation is the safe default:
- people wait to be told which tools to use, and how, rather than experimenting;
- AI mistakes are hidden or quietly worked around instead of reported, so the organisation never learns where the tools fail;
- decisions that AI should speed up still travel upward, so the promised gains in speed do not arrive;
- and the middle managers who would have to lead the change are the very people who feel least safe doing so.
An organisation in this position can buy and deploy AI tools and still not adopt them. On paper the tools are rolled out; in practice behaviour does not change, because the culture underneath has not changed.
Where findings like these point
The organisation chose not to continue beyond the diagnostic, so there are no results to report. Reading the findings today, they point away from more skills training and towards the conditions that make silence safe: clear decision rights at each level, so that taking a decision stops being a personal risk; shared measures across functions; and senior leaders who visibly reward challenge rather than only invite it.
The lesson for other leaders
How safe your culture feels from the top can tell you very little about how safe it is for the people who report to you. Measuring it is the reliable way to find out, ideally before you ask those people to lead something as uncertain as an AI transition.
Want to know how your own organisation looks from the inside? A Team Safety Check or AI readiness assessment is a good place to start, or book a diagnostic call.