First Principles
Rolling out the tool is the easy part. Knowing whether people are comfortable is the hard one.
Joe Dunleavy, Regional CTO for Europe and Global Head of Dava.X AI at Endava, on the twenty-year engineer the organisation doesn’t see, and the one thing about his own team he can’t get at by asking.
8 min watch
Joe Dunleavy
Regional CTO Europe & Global Head of Dava.X AI, Endava
Leads technology for Endava across Europe and heads the group’s global AI practice. His teams are distributed across continents, which means most of what he notices about them, he notices through a screen.
What this conversation is about
Joe is the first guest on this series to answer the opening question without hedging. He named a twenty-year engineer on his team — quiet, pivotal, and not widely known across the organisation. Then he made a distinction most leaders skip: under the radar is under the radar, so he isn’t under the radar to me. Invisible to the org and invisible to your manager are two different conditions, and only one of them is a career problem.
His early-warning signal isn’t in a tool. It’s a change in how someone carries themselves on a call — the person who always offers something up and suddenly doesn’t. He reads it over Teams and Zoom, across a team spread around the world, and he’s candid that it doesn’t always work. When a hard problem lands, he goes to his own boss, and every reason he gives is about temperament rather than expertise. Calm. Experienced. No judgment in the relationship. When something is already difficult, he says, you don’t take it to someone who adds more angst.
On AI he’s an optimist, and a specific one: more capacity, more opportunity, value creation rather than value capture. Then the final question turns the conversation. The one thing he’d want to see isn’t a number. It’s whether people have actually made peace with the technology — and he says, plainly, that he isn’t sure the answer he’d get by asking would be the true one.
Three things to take from this
1.
Invisible to the org and invisible to your manager are different problems.
Joe separated them without being asked, and the distinction is worth borrowing. If your manager sees you and the wider organisation doesn’t, that’s a distribution problem and it’s largely your manager’s to solve. If neither sees you, that’s a different and more urgent situation. Most people who feel overlooked have never worked out which one they’re in, and the two need opposite responses.
2.
Pick your crisis person for their nervous system, not their résumé.
Every criterion Joe gave was about state: calm, no judgment, doesn’t add angst. Not one was about domain expertise, and he’d been asked about a technical problem. That’s not a soft preference. Under real pressure your own capacity to think narrows, and a second person who is steady restores some of it, while a second person who is agitated takes more of it away. The most qualified person in the room is sometimes the worst one to call.
3.
You cannot be both the safe person and the person who ranks them.
Joe gets his honest conversation from someone with no stake in evaluating him. His team’s nearest equivalent is Joe, who does evaluate them. That’s not a flaw in how he leads; it’s structural, and it applies to every manager reading this. It means the most useful information about your team is the information least likely to reach you through a direct question, and building a better survey does not fix it.
“There’s no judgment in that relationship, which I think is important.”
— Joe Dunleavy
Host’s note — Dharma Ramasamy
Joe was asked who he turns to on a hard technical problem and answered entirely in terms of temperament. Calm. Experienced. No judgment. Nothing about skill, nothing about domain. I’ve asked that question enough times now to know how unusual it is to hear it answered that way, and how right it is.
There’s a reason the calm person is worth more in a crisis than the clever one. Threat is metabolically expensive, and the presence of someone trusted who has no stake in judging you lowers the cost of carrying it — the alarm still fires, it just stops having to be borne alone. That’s the mechanism underneath what everyone calls a good working relationship. The judgment-free part isn’t courtesy. It’s the active ingredient. Social evaluation is one of the most reliable ways to hold a stress response open, and removing it is what lets someone think again.
Which is where the asymmetry sits, and Joe walked right up to it. The same system runs in reverse when the other person holds your appraisal. Then the act of disclosing becomes the exposure, and the safest available answer is the confident one. So the more senior the person asking, the less true the answer tends to be — not through dishonesty, but because a nervous system in front of someone who can rank it does what it evolved to do. Joe has a relationship where that pressure is absent, and he named it as the thing he relies on. His team’s nearest equivalent is him, and he is also the one who assesses them. That’s not a criticism of how he leads. It’s true of every manager, and almost none of them have said it out loud.
The practical consequence is uncomfortable. The signal a leader most wants — how are my people actually holding up — is the one that stops existing the moment they’re the one asking for it. It has to be read from behaviour that nobody is performing.
Be on the series.
Five more of these this quarter. Twenty minutes, two cameras, nothing to prepare, and you see the edit before it publishes.
Get in touch →The thing Joe couldn’t ask for.
CultureGuard reads existing collaboration metadata to surface where human capacity is actually draining — before delivery slips, and without asking anyone a question they have a reason to answer carefully. No surveys, no AI, no content ever read.
How it works →