Leading 500: When Scale Breaks Decision Quality
The Org Chart Lied to Me In 2017, I ran a division that had scaled faster than anyone planned. Headcount grew from a manageable team into something requiring its own organisational chart – the kind printed on A3 paper and still needing a magnifying glass. By most visible measures, things went well. Revenue targets were hit. The division performed. And I, apparently, led it. Except I kept getting ambushed in ExCo. Not by crises, exactly. By facts. Things that happened two, three, four weeks earlier that I only learned about at the worst possible moment – across a table from people who clearly assumed I had known. The pattern was consistent enough to embarrass me, and consistent enough to force a reckoning. — The Situation At first, I assumed someone was dropping the ball. A manager not escalating. A process failing. A gap in accountability. So I did what most executives do when suspecting a human problem – I looked for the human. I found none. What I found was more uncomfortable. My team wasn’t hiding information. They were curating it. Every layer of management between me and the actual work made thousands of small editorial decisions weekly. What to escalate. What to resolve themselves. What to flag. What was “not worth the MD’s time.” These weren’t acts of concealment – they were acts of care, professionalism, even competence. A middle manager who escalates every minor issue to the divisional head isn’t doing their job. I understood that. I’d told people that. The problem was structural, not personal. I’d built a reporting structure. I hadn’t built an information structure. The distinction sounds small. It cost me dearly to understand it. The moment I remember most clearly was a Wednesday afternoon in late 2017. We were thirty minutes into an ExCo session when a colleague raised a client situation I’d genuinely never heard of. Not a rumour of – never heard of. I held my expression with the kind of carefully practised neutrality senior executives develop for moments like this. After the meeting, I spent forty-five minutes tracing the thread. The issue had originated six weeks earlier. It had been managed well at the operational level. But at no point in six weeks had anyone considered I might need to know it existed. They’d handled it. Case closed. They were right. They’d handled it. But I’d walked into an ExCo meeting carrying a version of reality six weeks out of date, and no one – including me – had known. — The Analysis First, I realised information loss is proportional to management layers, not individual competence. You can have exceptional people at every level, and you’ll still lose signal. Each layer is a compression. Each compression is a choice. Each choice is made by someone who isn’t you, under pressures you may not fully see, with a read on what you need that’s necessarily incomplete. This isn’t a performance problem. It’s a physics problem. The second insight was harder to accept: I’d trained my team to protect me from complexity. Not through explicit instruction. Through signals that accumulate quietly over years – expressions of impatience at detail, cultural norms about what gets raised in a room, an organisation that visibly rewarded resolution over escalation. My team were good pattern-readers. They read me accurately. I just hadn’t understood what signal I was broadcasting. The third thing – one I find most executives resist – is that the solution isn’t structural in the conventional sense. Flattening the hierarchy is usually the wrong answer. It creates other costs. What needed redesigning wasn’t who reported to whom. It was what got to me, in what form, and through what channel – regardless of the formal reporting line. Not all information should travel the same path. Not all decisions warrant the same proximity. Once I separated those questions from the org chart entirely, the problem became tractable. What replaced the old model wasn’t elegant or immediate. It involved explicit conversations about categories – the things I needed to hear about within 24 hours, the things I needed to see weekly regardless of status, the things I specifically didn’t need and was consciously delegating. It involved awkward recalibration of what “escalation” meant in our culture. It involved, frankly, a few direct conversations with people who’d been making entirely rational choices based on an implicit brief I’d never actually given them. — The Implication If you lead a large team and feel well-informed, examine that feeling carefully. The sense of being in command is often accurate. Occasionally, it’s a by-product of surrounding yourself, unintentionally, with people who are very good at making things look resolved. The organisations I’ve seen fail most quietly aren’t the ones with bad information. They’re the ones where the information was perfectly fine, but the wrong people decided what the executive needed to know. That decision – what reaches the top – is one of the most consequential in any organisation, and it’s almost never explicitly designed. It just happens. Mostly it happens well. When it doesn’t, it tends to matter at the worst possible moment.