There's a comfortable story about the arc of a technical career that goes: you build for a while, you get good at it, and then you graduate to advising — as if advice were a higher altitude you ascend to once you've paid your building dues, leaving the keyboard behind like a childhood bicycle. I've been doing more advisory work lately alongside the hands-on delivery, and I want to complicate that story, because it's wrong in a way that matters. Advising isn't above building. It's a different job that depends utterly on the building — and the trick is knowing precisely which of your building instincts to carry across intact and which to leave at the door, because a few of them will actively sabotage you.

What transfers — and is the whole point

Start with what carries over, because it's the part the "graduation" story gets right for the wrong reasons. The single most valuable thing I bring to an advisory conversation is having actually built the thing I'm advising on. That's not nostalgia; it's the entire source of the credibility.

  • Knowing what breaks at 2am. When I tell someone a particular architecture will hurt, it's not from a slide — it's from having been paged by that exact architecture. That specific, scar-tissue knowledge is what separates advice worth paying for from advice you could get from a whitepaper, and it comes only from having built.
  • Smelling the gap between the demo and the reality. Years of watching features that dazzle in a keynote wilt under a real workload gives you a nose for it. In advisory work that nose is the product: I can tell a client which of the shiny things will survive contact with their actual data, because I've watched the equivalents survive or die in mine.
  • Speaking the builder's language. The engineers on the other side of the table can tell within minutes whether an advisor has ever actually done the work. Having built means I can talk to them as a peer, which means they'll tell me the truth about the state of their systems instead of the sanitised version they'd give a suit. That access is worth everything, and it's non-transferable — you either earned it at the keyboard or you didn't.

This is why I hold, firmly, that the hands-on years aren't a phase you leave behind. They're the asset. Across a dozen-plus organisations the pattern only ever became legible because I'd had my hands in the specifics first. Advice without that underneath is just confident generality, and confident generality is exactly what my clients have too much of already.

What doesn't transfer — and quietly sabotages you

Now the harder half, the part the graduation story misses entirely: some of what makes you an excellent builder makes you a worse advisor, and you have to consciously put it down.

  • The urge to solve it yourself. A builder's reflex, faced with a broken thing, is to fix it — to reach for the keyboard and make it right. In an advisory seat that reflex is a trap. My job there isn't to solve their problem for them; it's to help them become able to solve it, and every time I grab the wheel I've made them more dependent and taught them nothing. Learning to sit on my hands while someone works toward an answer I can see faster has been genuinely hard.
  • Depth for its own sake. Building rewards going deep — the elegant detail, the clever optimisation, the last 5%. Advising punishes it. The client with a strategic decision to make does not need my beautiful nuance; they need the shape of the answer at the altitude they're deciding at. My instinct to add precision usually just adds fog. Knowing which 90% of what I know to leave out is a skill I did not have as a builder and had to develop from scratch.
  • Loving the tool. When you build with something daily you develop affection for it, and affection is bias. An advisor's value depends on not being in love with any particular tool, so a client gets an honest answer rather than my favourite one. I've had to deliberately unlearn some of the fondness that made me a productive builder, because the moment a client senses I'm recommending what I enjoy rather than what they need, the advice is worthless.

The synthesis: an advisor who never stopped building

So where does that leave the arc? Not at "build, then graduate to advising." For me it's landed at something more awkward and more honest: keep doing both, because each protects the other. The building keeps the advice grounded in what's actually true this year rather than what was true when I last touched a keyboard — advice has a shelf life, and hands-on work is how you refresh it. And the advising keeps the building pointed at what actually matters rather than at interesting-for-its-own-sake.

The best data advice doesn't come from people who've left building behind. It comes from people who never fully did — close enough to the keyboard that the advice is still true, far enough back to see the shape. Lose either distance and the advice goes off.

The danger in the "graduation" framing is that it quietly tells you the building was the ladder and advising is the roof — that you can pull the ladder up once you're on top. Do that, and your advice slowly detaches from reality, aging into the generic strategic voice that's fluent about everything and grounded in nothing, the kind that clients have learned to nod at and ignore. What I'm trying to protect against is exactly that drift. The hands-on work isn't the thing I did before the real job. Increasingly, it's the thing that keeps the real job honest — and the day I can't credibly say "I built one of these and here's what actually happened," my advice is worth a good deal less, whatever the title on the invoice says.