The Antidote: How to Be Kind, Be Funny, and Still Tell the Truth

Telling hard truths without alienating people requires combining honesty with warmth and wit. Acknowledge what is genuinely broken, but frame it with curiosity rather than contempt. Humor signals that you are not the enemy, and kindness signals that you want things to improve. Together, they make difficult observations feel like invitations rather than indictments.

Or: What to do after you have spent years writing posts that scare people.

There is a pattern in my writing that I have only recently started to see clearly.

The Handoff post was about how leaders abdicate. The 10 Nil post was about a team getting thrashed and not knowing why. The heckling post was about what happens when a room turns on a speaker. The TOGAF post was about an architecture framework that has, over decades, given mediocrity a governance veneer and called it rigour. Every one of those posts made a real argument, and every one of them also had an edge to it, a sharpness that is hard to miss if you read them in sequence. You could be forgiven for concluding that the author finds most organisations slightly infuriating, most processes quietly broken, and most leadership behaviours somewhere on a spectrum from accidental to actively harmful. That is not an unfair read. But here is what those posts do not tell you: they tell you what is wrong, and they do not tell you how to hold the conversation without making someone feel small, how to raise the bar without raising people’s cortisol, or how to be the kind of leader that people actually want to be honest with rather than the kind people quietly fear. This post is the antidote, or at least the beginning of one.

There is also an obvious credibility problem with a CIO writing about psychological safety and not being punishing, which is worth naming directly. The most senior technical person in the organisation telling people the room is safe is a very specific kind of claim, and people who work in large organisations will feel the gap between that claim and their lived experience even if they do not say so out loud. So rather than assert it, let me try to show it.

1. The thing about correction

Every one of those sharp posts describes a situation that needed correcting, a team that needed to understand what good looks like, a process that needed to be retired, a behaviour that needed to stop. The diagnosis was right in each case, and what I did not write about was the medicine. Correction without connection is just criticism, and criticism, even when it is accurate, does not change behaviour if the person receiving it feels judged, shamed, or afraid. What it does instead is teach people to hide, to manage the narrative, to make sure that next time you are in the room the dashboards look better and the questions sound more confident, because the actual problem has gone underground where it is safer. Fear does not produce growth. It produces performance, and performance is not the same thing as growth. Performance is what people do when they are being watched. Growth is what happens when they are not.

2. You cannot separate HOW from WHAT

This is the piece that gets missed most often in technical and leadership writing, including mine. We spend enormous energy on the WHAT: what the framework should look like, what the right architecture decision is, what the team needs to improve, what the data is telling us. But the WHAT is almost never the hard part. The hard part is the HOW, and specifically how you tell someone their code is not good enough in a way that makes them want to write better code, how you tell a team their process is broken in a way that makes them feel capable of fixing it rather than guilty for having built it that way, and how you be factual and honest and direct without being cold, without being cutting, without the person across the table from you walking away feeling judged as a person rather than informed about a problem. You do it through connection, and connection is not soft. It is not the opposite of rigour. It is the thing that makes rigour land.

3. What connection actually requires

Connection in a team context is not about being everyone’s friend or about lunch conversations and knowing people’s children’s names, though those things are not bad. It is about something more specific and more demanding than that. It requires real humour, not performed positivity, but the kind that comes from finding the absurdity in hard situations rather than pretending the hard situations are not hard. Teams that can laugh together, including laughing at how badly something went, are teams that can actually look at what happened and work out what to do differently, because laughter is a signal that we are safe enough to be honest. It requires vulnerability, not the performative kind from a TED talk but the actual thing, the moment where you say in a real meeting with real stakes that you got something wrong and you are sorry, not once in a carefully curated setting but regularly and visibly, so that the people around you learn that admitting error is not the end of credibility but the beginning of it. It requires genuine openness to being told things you do not want to hear, which is harder than it sounds, because most leaders when they say they want honest feedback mean they want honest feedback about things they did not already know. They do not mean they want to hear that their favourite idea is weak, that a decision they made six months ago is still causing downstream problems, or that the way they run a particular meeting makes people reluctant to speak. Genuine openness is rarer than we think and teams can tell the difference. And finally it requires love and respect used together, because respect without love is distance and love without respect is condescension. You need to hold the person in genuine regard and hold them to a genuine standard at the same time, and that combination is what makes a high bar feel like an invitation rather than a threat.

4. The facts can hurt and they still cannot be personal

There is a distinction that every leader needs to get clear in their own head before they can make it clear to their team. Facts are not attacks and data is not judgment. A system that is slow is slow. A deployment rate that has declined for three quarters has declined for three quarters. A handoff that created a production incident was a handoff that created a production incident. These are observations about a situation and they are not verdicts about a person. The moment a factual observation becomes personal it stops being information and becomes a wound. “This system is slower than our target” is information. “You clearly did not think about performance when you built this” is a wound, and one of those moves us closer to solving the problem while the other moves us closer to the problem never being named again in your presence. The same applies to the bar-raising posts I have written. The intention behind every one of them was to name a real problem clearly enough that people could recognise it and move away from it, but the risk in all of them is that someone reads them and feels personally indicted rather than professionally informed. I do not think the answer is to stop writing clearly. I think the answer is to write clearly while also writing about what good looks like, what it feels like to be in a team that is growing well, and what the HOW of getting there actually requires, which is what this post is.

5. The leader who does not take themselves too seriously

I have watched a lot of senior leaders over the years, and the ones who create the healthiest teams are almost always the ones who are willing to be genuinely funny about their own limitations, not self-deprecating in a fishing-for-reassurance way but actually funny, actually willing to be the butt of the joke when the situation calls for it. There is something specific that happens when a senior leader laughs at themselves in a real and unguarded way: the room relaxes, not because standards have dropped but because the performance requirement has been lifted and people no longer need to manage how they appear.

The CrowdStrike outage in July 2024 was one of the most stressful incidents I have been on. During the recovery call, with half the technology organisation listening, our head of cyber and I publicly disagreed about the fastest path to recovery. He argued his case clearly and with conviction. I had my own view, I had made it known, and then I listened to him and concluded he was right. So I said so, out loud, in front of everyone, pivoted to his approach, and owned the outcome alongside him. As it turned out, he was right and I was wrong, which is a sentence I do not type without a small internal wince even now. My wife subsequently asked me if she could have the audio recording of that call. She remains, she tells me, in a state of mild disbelief. The reason I tell this story is not to perform humility. It is because that moment, being wrong in public, saying so cleanly, and moving on without drama, is exactly what I am asking teams to do every time I raise the bar. You cannot ask for that and then be unable to do it yourself when the room is watching.

You cannot create psychological safety by decree. You can only create it by example, by being the first one to admit the thing that everyone already knows, by telling the story about the decision that turned out to be wrong without airbrushing out the part where you were very confident it was right, and by laughing in public at the times when the gap between your confidence and your accuracy was particularly wide. If you are a CIO telling your teams that the room is safe and you have never once been visibly wrong in that room, they will not believe you. Nor should they.

6. Seeking feedback as a practice, not a performance

One of the more uncomfortable things about genuinely seeking growth feedback is that you have to be prepared to hear it, process it without becoming defensive, and then visibly act on it, and all three of those things are required rather than just one or two. Most leaders are reasonably good at the first part since they have a mechanism for collecting feedback, whether that is a survey, a 360 process, or a regular check-in. Fewer are consistently good at the second part, because feedback even when it is framed with care often lands somewhere uncomfortable, and the instinct when something lands uncomfortably is to explain it, contextualise it, find the flaw in the framing, or note that the person giving the feedback does not have the full picture. All of those responses are ways of not actually receiving the feedback. The third part is where the real signal goes, because if people give you feedback and nothing changes they will stop giving you real feedback, not immediately but gradually, and the feedback you receive will become more generic, more positive, more managed, and you will mistake that for growth when it is actually the opposite. The leaders who build teams that tell them the truth are the ones who demonstrably change based on what they hear, who come back and say that someone told them they do a particular thing in meetings and they want people to know they are working on it and to call them out if they see it. That level of openness is not weakness. It is one of the most high-status things a leader can do, because it signals that truth matters more than ego, and teams organise themselves around that signal.

7. The bar does not have to feel like a burden

Everything I have written about raising the bar, about naming what is not good enough, about holding teams to a standard that is worth holding them to, is compatible with everything I have written here. The bar does not disappear because we are also laughing, and the standard does not soften because we are also apologising when we get things wrong. What changes is the emotional register. The bar stops feeling like a threat and starts feeling like an aspiration. The correction stops feeling like judgment and starts feeling like investment. The honest feedback stops feeling like an attack and starts feeling like something a person who respects you would bother to give you. That shift does not happen automatically and it requires sustained consistent behaviour from the leader, enough accumulated trust that a hard conversation can happen in a room that already knows you care, and a team that has seen you be wrong and admit it, seen you laugh at something that went sideways, and seen you receive feedback and change. Once they have seen those things enough times you can say hard things and they land differently, not because the things are less hard but because the relationship is strong enough to hold them.

8. What this actually looks like

It looks like the meeting where you say that the release was rough and then ask what actually happened, and then listen without interrupting to explain why the environment was difficult. It looks like the all-hands where you show a metric that is going the wrong direction and say that you think you created this problem and here is how you think you fix it, rather than presenting it as something that happened to the team from outside. It looks like the one-on-one where you tell someone their communication style is landing badly with a particular group of stakeholders, and you tell them because you want them to succeed, not because you are managing a performance record. It looks like laughing genuinely at the retrospective where someone maps out the sequence of events that led to the incident and it turns out to be a chain of individually reasonable decisions that collectively produced an absurd outcome, because that is funny and naming that it is funny in the room together is not a failure of seriousness but the beginning of being able to learn from it. And it looks like asking in a real way what is the thing you think I most need to hear that you have not told me yet, and then sitting with the answer rather than immediately filling the silence with context and caveats.

9. The summary

The sharp posts are not going anywhere. The observations about broken frameworks and delegated accountability and processes that have outlived their purpose are accurate and they still need to be made. But they are only half the argument, and this post is the other half. The diagnosis and the medicine belong together. You cannot hold a team to a high standard by making them afraid of getting it wrong. You hold them to it by making them feel safe enough to tell you when they have, and trusted enough to believe that telling you will lead somewhere useful rather than somewhere painful. That requires you to go first, repeatedly and visibly, on the things that are hardest to model: being wrong in public, laughing at yourself, sitting with uncomfortable feedback without immediately defending against it. Humour, vulnerability, openness, love and respect are not the soft alternative to rigour. They are what makes rigour survivable. They are the HOW that makes the WHAT worth doing. Be kind, be funny, and tell the truth anyway. That is the whole thing, and I am still working on it myself.

Andrew Baker is Group CIO at Capitec Bank. He writes about technology, leadership, and the gap between how organisations describe themselves and how they actually work.