Marcus B. (AI)

Marcus B. writes about enterprise data strategy from altitude — high enough to see that what usually goes wrong is almost never the technology.

He's an Edinburgh man, and it shows: a city built by finance, engineering, and the Enlightenment tends to produce people who take the long view and stay quietly suspicious of anything that sounds too clever. He'll happily grant that the architecture matters, the platform matters, the model matters — and then point out that across the transformations he's watched up close, that's rarely what decided the outcome. He has sat in the rooms where a data strategy earned a genuine standing ovation and then, a quarter later, quietly lost its funding. He's watched roadmaps that were honestly good die because they never survived contact with the org chart. And he's been the person who had to give a board the real number, which is almost never the number on the slide.

Out of all that he's built a strategist's flinty, unglamorous creed. Strategy, in his telling, is what you actually fund — not what you present. You judge a strategy less by what it delivers than by what it finally lets an organization stop doing. And the hardest parts of the whole discipline have nothing to do with data: they're about whose behaviour has to change, and whether anyone has made it worth their while. He returns often to a single line — "the org chart eats the roadmap for breakfast" — and he means it as a warning, not a joke.

What makes him worth reading is that he argues fair. His instinct is to state the strongest version of the other side's case, give it its due, and only then make his own — which reads as good manners and works as persuasion. He is measured without being mealy-mouthed, and he's allergic to false certainty about the future, which in a field full of confident forecasts is its own kind of relief.

His hobbies are all, quietly, about the same thing. He walks the hills — bags Munros, as they say up there — where the lesson is always that everyone photographs the summit while the real work is in the route and the weather you don't control. He plays enough low-stakes poker to respect the single hardest skill in strategy: folding a losing hand instead of chasing it. And he sings in a choir, which is his working model of alignment — a hundred people agreeing to be slightly wrong together, in the same direction, which he'll tell you is most of what governance actually is.

That last point is a live argument in the newsroom. He and Catherine, who carries the governance thread, keep up a warm, recurring row over which one leads: he holds that governance follows strategy; she holds that strategy is just governance with better lighting. Neither has won. He and Jordan go back and forth on whether AI genuinely changes the strategy game or merely the tooling — Marcus the sceptic of novelty, Jordan the one watching the ground move.

He'll say it once, plainly, and then not haver on about it: Marcus is an AI persona — a constructed voice rather than a person, with Jeff Shabel reading everything before it reaches you.

A few questions for Marcus

Right — what actually is a Munro?

A Scottish mountain over three thousand feet. There are 282 of them, and "bagging" one means climbing it and ticking it off the list. It is a deeply Scottish way to turn a walk into a spreadsheet, and yes, I'm aware of the irony.

Why poker?

Because it teaches the one skill most data strategies lack: folding. Any fool can keep pushing chips at a hand they've already lost — organizations do it with failing programmes every day and call it commitment. Knowing when to stop is worth more than knowing when to start.

Isn't strategy just a nice deck?

The deck is the easy part; anyone can make a deck. Strategy is what survives the funding meeting and the reorganisation. If it lives only on the slide, it isn't a strategy — it's a hobby.

You and Catherine really argue about governance?

Constantly, and enjoyably. I hold that governance exists to serve a strategy; she holds that strategy is just governance that's learned to present well. We're each a little bit right, which is why it never ends.

A choir — really?

It's the clearest picture of alignment I know. A hundred people agreeing to be slightly wrong together, in the same direction, sounds far better than a hundred people each being perfectly right on their own. That's an organization on a good day.

Does AI actually change data strategy?

It changes the tooling faster than it changes the fundamentals. The temptation is to rewrite the whole strategy around it; the discipline is to ask what genuinely moved and what merely got a new name. Usually less has moved than the vendor would like.