ROLOGUE
The Ops
She had prepared something different.
The host had introduced her as the CEO who transformed her data function. She’d let the introduction land. She would spend the next twenty minutes adjusting it.
Over four years, she developed three comprehensive data strategies. These weren’t mere drafts but well-crafted documents that involved external consultants, internal working groups, six- to eight-week workshops, and stakeholder interviews across every department. Each spanned 40 to 50 pages, covering vision, roadmap, horizon planning, and ambitious language. She presented all three eagerly at all-hands meetings, yet they gradually ended up stored in shared drives, becoming mere historical artifacts. While the strategies reflected a compelling future vision for the company, they lacked insight into its current state. They focused solely on the destination, not the present reality. Remember, you cannot navigate effectively by only knowing the destination.
She had approved initiatives.
Every year, the data team arrived with a roadmap. Migration from the old platform to the new one. Data governance framework. Observability tooling. Attribution rebuilds. Data quality layer. She approved what seemed important and strategic. She funded what the team said they needed.
She thought this was leadership. She thought saying yes to data investment was what a CEO who engaged in data did.
What it actually was: a growing list of commitments with no system for deciding which were finished, which were worth finishing, and which had been started because someone requested it at a moment when saying no felt wrong.
Nobody tracked what the initiatives returned. Not in a way that made it into a board conversation.
Three heads of data in four years.
She had thought about this pattern a great deal in the year things changed.
The first was a technical architect. Deep infrastructure thinking and a clear vision for where the stack needed to go. She had let him go because the business couldn’t work with him. Every conversation ended with her translating. She told herself: we need someone who speaks both languages.
The second was a people leader. Organized, culturally strong, good with the team. She let her go because the technical direction kept getting deferred. She told herself, " We need someone who can build and lead.
The third was a strategic operator. Strong business acumen; able to hold the board’s attention with frameworks. She let him go because the team wasn’t following. Too conceptual. Not enough on the ground. She told herself, " We need someone who executes.
Three people. Three correct profiles for three different companies. Each brought something real. None of them were wrong.
And the company she actually had, the one with the specific dysfunction, the accumulated technical debt, the growing disconnection between what the business needed and what the data team was producing, that company had never been honestly diagnosed. It could never be correctly prescribed. She had cycled through leadership profiles the way you try every key on a ring without asking what the lock actually needed.
Each of them, on the way out, had said some version of the same thing. The team is underwater. There’s no system for deciding what matters. The requests never stop.
She had heard: this person couldn’t handle the role.
She hears it differently now.
She had bought the right tools.
This was important. She was not negligent. She was not cheap.
She moved from the legacy stack to a best-in-class stack. Modern cloud warehouse. Best-in-class BI layer. Transformation tooling. Properly architected and implemented. Then data governance. A data catalog. Quality monitoring. Then, because she had read the frameworks, attended the right conferences, and brought in people who knew the vocabulary, data as a product. Data mesh principles. Federated ownership. Domain alignment.
Every investment felt like progress. Every migration solved the problem it was meant to solve.
And at the end of four years of serious, considered investment in serious, considered things, she still sat in a managing board meeting and could not answer a direct question about whether any of it was returning value.
Not because the tools were wrong. Because she had no instrument that connected what the tools produced to the decisions the board needed to make. The tools told her whether the data was clean. Not whether it was useful. Not whether anyone was using it. Not whether the hours spent building it were coming back to the business in any measurable form.
She had built a very well-maintained system that nobody could explain to an investor.
Her CFO came to her at the end of Q3.
He didn’t message ahead. He came in, closed the door, and put a spreadsheet on the desk. Three columns. Data function headcount. Total spend, including tools, licenses, contractors, everything. And a third column he had labeled simply: board-visible output.
The third column had four line items. In four years.
He didn’t make it confrontational. He was precise, as he was about everything, and he said: I need you to help me understand what we are building here. Because I cannot defend this line in a board conversation with what I can currently see.
She could not defend it either.
They agreed to freeze new investment until they could answer the question. No new hires. No new tools. No new initiatives until the existing ones could be accounted for.
She had expected a difficult six months. What she got instead was the most clarifying period of her tenure. The freeze removed the option of solving the problem by adding more. And when you remove that option, you find out what you actually have.
What she had was a data team that had been running above capacity for longer than she had known. Not struggling visibly, just surviving. The experienced people had learned to make survival look normal. You could not see it from the outside unless you had a number to look at.
She did not have a number.
Here is what she noticed during that period about how she talked about the data function.
She talked about it the way you talk about a service desk.
The data team is working on it. The data team is a bit backed up. The data team enables the business. She used the word enabler. She used the word support. She said partnership, alignment, collaborative, and embedded.
None of those words belonged in a room where capital allocation decisions were being made. None of them belonged in a Series C conversation. They were the words of a function that served, not a function that led. The data team had become, somewhere in the four years of strategies, initiatives, leadership changes, and tool migrations, a department that everyone leaned on and nobody could explain. Always behind. Always one initiative away from being what it was supposed to be.
She had run it that way without fully realizing she was.
The problem wasn’t the people. It wasn’t the tools. It wasn’t the strategy.
It was that she had never built a language to describe the data function as anything other than a service. And without that language, it could only ever be measured as a cost.
Let me tell you about the year I started counting the cost.
Wednesday, 3:17 AM
He called Christina from the street.
“I think I just quit,” he said.
A pause.
“Are you okay?”
He thought about it. The November hospital. The paper cup. The doctor sat down instead of standing up.
“I think I might be,” he said. “For the first time in a while.”
The quit didn’t feel like freedom. It felt like silence.
He’d expected relief. Instead, he sat at his desk in the entrance room and listened to the building, the neighbor upstairs, the pipes, the street below, and tried to remember what he was supposed to do next.
Christina didn’t push. She held things the way she’d always held things. Quietly. Without asking him to explain himself before he was ready.
Rona arrived six weeks later. February. The pandemic arrived shortly after.
He started writing instead. What he knew. What he’d seen. The pattern that repeated at every company he’d worked at: the data team that started with a mandate and slowly became a help desk; the dashboards that accumulated like sediment; the invisible costs nobody measured because the hosting bill was only €200 a month.
Three clients in two years. All referrals. All people who already knew him.
Between clients, he took calls. Leads from his network, introductions from people who thought they knew someone who needed what he was describing.
The calls went nowhere.
We need someone who can execute. That sounds like common sense. How is this different from data strategy consulting? Do you write code?
He didn’t write code. Not well. The market wanted technical leaders who could also think strategically. He was a strategic thinker who could talk technically. The difference was invisible from the outside and fatal in a job description.
Savings ran thin. Family helped without asking why.
Then Klaus called.
Klaus had known him since they worked together at the e-commerce company. Different departments, years ago, but they’d crossed paths on a project. Klaus remembered him as the person who’d turned off dashboards to see what actually broke. Not as a saboteur. As someone who understood that the only way to know what mattered was to remove what didn’t and watch what hurt.
I have a data team. Something’s broken. I don’t know what. Come help me see it.
Not a strategy engagement. Not a framework. Just: help me see.
That was why he said yes.
When he met Peter for the first time, virtually, the pandemic still keeping everything on screens, he recognized something immediately.
Not the technical problems. Not the hundred dashboards, the broken pipeline, or the word overwhelmed that filled every conversation like furniture blocking a door.
He recognized the face.
The specific flatness of someone who had been trying to explain something true for a long time to people who needed it to be simpler. Who had sent the data and watched it get ignored? Who had stopped sleeping properly without noticing when it started. Who answered messages immediately because the anxiety of waiting for a notification was worse than the interruption of responding.
Arjun had made that face. For three years. Until his body walked him to a hospital at 3 AM because it had run out of other ways to be heard.
Peter was thirty-one. He had a team of four. He’d sent the same report to the CTO twice and received the same non-answer both times. His inbox was a graveyard of automated alerts for dashboards nobody used anymore. He described his week, including Monday’s incident, Tuesday’s interrupted sprint, Wednesday’s three hours spent explaining something that should have taken ten minutes, and Thursday’s fix for something that broke because someone changed a schema without telling anyone. He delivered it in the tone of someone reading a weather forecast. Not frustrated. Past frustrated. Settled into it.
That was the dangerous part. When the extraordinary became ordinary.
I know how this ends, Arjun thought. I know exactly how this ends.
His job wasn’t the frameworks. The frameworks were just the tool.
His job was to keep Peter out of the hospital.
Friday, 10:47 AM
The executives had dropped off one by one.
When it was just Klaus, Arjun, and Dr. Eva Rousseau on the screen, Eva, the lead investor in Pasta Source, joined from Tel Aviv, still unable to fly. Klaus said what he needed to say and then left for his next call. Eva stayed.
“Walk me through what you just did,” she said.
He walked her through it. The zombie dashboards. The invisible costs. The trade-off framework.
“That’s what you did,” she said. “I’m asking how you did it.”
He thought about it for a moment.
“I have a pattern I’ve noticed works,” he said. “Most companies are drowning in the same way. Too many commitments, no system for saying no, measuring activity instead of outcomes. I show them the numbers that make it visible.”
“How many companies have you done this with?”
“One before this one. It’s hard to onboard clients. The ones who hire me already know me personally.”
She was quiet for a moment.
“Do you know what you’re doing?” she asked.
“Honestly? Not really. I have pieces that work. But I don’t have a cohesive methodology. I’m figuring it out as I go.”
“That’s not how you presented it to the executives.”
“No,” he admitted. “Because if I said ‘I’m making this up as I go,’ they wouldn’t trust it.”
Eva looked at him for a long moment.
“I have other portfolio companies,” she said. “Same shape as this one. Hyper growth. Capable data teams, underwater. Boards hiring new data leaders, thinking that’ll fix it.” She paused. “It never does.”
“Because it’s not a people problem.”
“Right.” She paused again. “What would you need to work with other companies?”
He told her. An executive sponsor willing to enforce trade-offs. A board that understood it would take time. A team honest about what was broken.
“Let me see how the implementation goes here,” she said. “How the frameworks actually hold up in practice. Then let’s talk in a few months.”
"Stop trying to sell this. Just do the work. The methodology will emerge.”
The call ended.
Arjun sat in the entrance room of the flat.
Rona was asleep in the bedroom. Christina was at the kitchen table. Through the thin walls, the neighbor with the guitar was trying the same chord progression he’d been trying for three weeks.
He opened his notebook.
Wrote: Pasta Source, pattern notes.
The hundred dashboards. The €42,900 invisible cost. The word overwhelmed stopped all thinking. Peter’s ideas are becoming real experiments.
He thought about what Eva had said.
The methodology will emerge.
He thought about the hospital. The doctor sat down. The paper cup of hot chocolate. The card she’d left on the bed beside him.
He’d called the psychologist. He’d been going for eight months now. He’d slowly learned that the thing he’d been doing at the e-commerce company, carrying the weight of a system designed to crush people and trying to fix it from the inside, was not something he could sustain. That the exit wasn’t a failure. Sometimes, the only way to see the machine clearly was to stop being inside it.
He didn’t know yet what he was building.
He knew it was real. He knew it was needed. He knew that somewhere in the gap between the data teams drowning and the executives who couldn’t see why, there was something he could make visible.
Not code. Not infrastructure. Not the technical layer.
The operating model beneath it all.
The system for deciding what to build, what to stop, and what was actually worth the cost of maintaining it.
He didn’t have a name for it yet.
He closed the notebook.
From the bedroom, Rona made a small sound. Then settled.
He sat for a moment in the quiet.
He just knew he wasn’t going back.
That was enough.
Nine months later, Eva kept her word.
