I've just passed DA-100, "Analyzing Data with Power BI," which earns the Data Analyst Associate certification. And I can already hear the reasonable question: I lead a data platform team, I'm an engineer and increasingly a manager, so why am I sitting an exam pitched at Power BI analysts — the people who consume the platform rather than build it? It's a fair question, and the answer is the whole point of this post: because the best way to build something for a role is to understand that role from the inside, and this cert deliberately put me in a chair I don't usually sit in.

Building for people whose job you've never done

Here's a trap I've watched platform teams fall into, and been guilty of myself: you build the platform, the pipelines, the models — all the engineering underneath — and you hand it to the analysts who'll actually turn it into insight, and you do so with only a vague, second-hand sense of what their day is actually like. You've optimised the plumbing without ever really living downstream of it.

And it shows. The platform that's beautifully engineered but subtly awful to analyse on top of is a real and common thing. The model that's technically pristine and yet makes the analyst's life miserable because it wasn't shaped for how they actually work. You can't see those failures from the engineering side; you can only see them from the analyst's seat, feeling the friction they feel.

So DA-100 was, for me, a deliberate exercise in sitting in that seat. Properly learning Power BI from the analyst's perspective — not "how do I connect a data source" as an engineer, but "how do I actually build reports, wrangle a model, wrestle DAX into shape, and turn data into something a business person understands" as the person who does that all day. I wanted to feel the downstream experience of the thing my team builds, so I could build it better.

What sitting in the analyst's chair taught me

The cert covers the analyst's whole workflow — preparing and modelling data, doing the analysis, building the visualisations, and managing the deliverables — and going through it with an engineer's eyes was genuinely illuminating. A few things landed:

  • The model is where the analyst lives or dies. I knew this abstractly. Doing the analyst's work, I felt it: a well-shaped model makes analysis flow and a badly-shaped one makes every question a fight, and the difference is entirely upstream, in decisions my side of the fence makes. It gave me a much sharper sense of my own responsibility for the analyst's experience.
  • DAX is a genuine skill, and respecting it changed how I think about supporting it. Really learning DAX — the calculation language behind Power BI — rather than waving at it from the engineering side, gave me respect for how much craft the analyst role demands, and ideas about how the platform could make that craft easier rather than harder.
  • The "last mile" is where value actually gets delivered or lost. All my platform work is upstream of the moment a human looks at a chart and understands something. That last mile — the analysis and the communication of it — is where the entire value of everything underneath is finally realised or squandered. Spending time there reminded me that the platform is a means, and the insight is the end.

The broader principle

There's a general lesson here I keep relearning, and it's not really about Power BI. To serve a role well, understand it from the inside. As I move further into leading — building platforms and teams for other people to do their work — the temptation is to optimise from my own perspective and assume I know what the people downstream need. The corrective is to actually go and be them, for a while, and feel what they feel.

It connects to the thing I believe most about this whole field: that the value is in the human on the other end, and you can't serve a human whose experience you've never inhabited. An engineer who's never analysed builds for an imagined analyst. An engineer who's sat the analyst's exam, felt the friction of a bad model, and wrestled the DAX builds for a real one.

The best way to build something for a role is to spend enough time in it to feel where it hurts. A certification aimed "below" your level isn't a step down — it's a deliberate trip downstream to see what you've been floating toward the people you serve.

Worth it, for an unusual reason

So was DA-100 worth it for a platform lead? Yes — but for a different reason than it's worth it for an aspiring analyst. For them it's a credential for a career step. For me it was empathy through practice: a structured way to inhabit the perspective of the people my work exists to serve, and to come back to my own side of the fence with a sharper sense of how to build for them.

I'd recommend the move to anyone leading a platform team: go and get certified in how the people downstream of you actually work. Not for the badge, but for the view from their chair — because you cannot build a good platform for a role you've only ever watched from a distance, and the fastest way to stop watching from a distance is to go and do the thing, exam and all.