It feels almost absurd to write a "year in data" reflection about 2020, a year that most of us would rather simply survive than analyse. But the ritual has its uses, and looking back over these twelve strange months, I think there's a genuine and coherent story about what happened to data work this year — one that isn't really about any new technology at all. Because 2020 didn't hand us a shiny new tool. It did something stranger and more consequential: it compressed years of gradual change into a few violent months, and it sent all of it home.

The year didn't invent anything — it accelerated everything

The temptation with a year-end review is to name the big new thing. But 2020's defining feature in data wasn't a launch. It was an acceleration. Every slow trend that had been creeping along for years got shoved forward at once, by a global emergency that removed all the usual reasons to defer.

  • The move to cloud stopped being optional. Organisations that had been "thinking about" cloud for years discovered, in a single week, that infrastructure sitting in an office they could no longer enter was a liability, and infrastructure in the cloud was the thing keeping them working. A gradual migration became an emergency necessity overnight. I felt this personally and gratefully — our work continued precisely because it was already in Azure, and I watched teams whose systems were trapped on-premises have a categorically worse year.
  • Digital transformation lost its quotation marks. The phrase had been a slightly hollow bit of consultant-speak for a decade. In 2020 it became simply what happened, because there was no alternative. Things that were "on the roadmap for next year" shipped in a fortnight, because the choice was ship it or stop.
  • Data literacy became suddenly, visibly urgent. A pandemic is, among many more important things, a data story — case curves, exponential growth, the difference between a good chart and a misleading one, live on every front page. Everyone became an amateur reader of data, well and badly, and the value of being able to read and communicate data honestly went from a nice professional skill to a matter of public consequence.

None of these was new. All of them were accelerated past the point of gradual, and the result is that data work enters 2021 several years further along than the calendar alone would suggest.

Everything moved home — including the data conversation

The other half of the story is the literal one: everything moved home. Work moved home. The office, the desk, the corridor, the whiteboard — home. And with them, subtly, the whole texture of how data work happens.

I've written already about what it's like to run a data team remotely — how the systems came with us easily because they were in the cloud, and how the human layer was the part that broke, the ambient sharing and the reading of the room that don't survive being moved to a screen. Stretched across a whole year, that shift has done something lasting. We've been forced to make explicit what used to be implicit — to write down the knowledge that lived in the air, to schedule the contact that used to just happen, to communicate deliberately because osmosis stopped working. A team whose collaboration was tacit has had to become a team whose collaboration is articulated.

And I suspect some of that is permanent, and some of it is even good. Being forced to write down what was only ever known, to communicate on purpose rather than by proximity, has made us more legible to ourselves and more robust. The office was quietly letting us avoid a lot of that discipline. Home wouldn't let us.

The year I became a manager, into all of this

For me personally, 2020 has an extra layer, because it's the year I stepped from building the platform to leading the team — and I did it into all of the above. I learned to manage a team remotely before I'd properly learned to manage a team at all, which is a strange order to do it in, and possibly a formative one. Everything I'm learning about leadership, I'm learning through a screen, in a year where the usual scaffolding of how you learn to lead — watching other leaders up close — was shut off.

I mention it not because my small career pivot matters against the scale of the year, but because I think a lot of people had a version of it: some significant personal or professional change, undertaken in the middle of a general upheaval, with none of the normal supports. 2020 made everyone improvise, and improvising is exhausting, and also, occasionally, a way you discover you can do more than you thought.

2020's legacy in data isn't a technology. It's a compression — a decade of slow migration, transformation, and the rising value of data literacy, all forced to happen at once, at home, with the tacit made explicit. We didn't get a new tool. We got shoved several years into the future.

What I'm carrying into 2021

I'm wary of tidy silver linings about a year this hard — too many people lost too much for "but we learned to work remotely!" to be anything but slightly obscene as a summary. So let me keep it modest. The technology story of 2020 is that the future arrived early and unevenly, and we're now living in a more cloud-based, more remote, more data-literate world than we were meant to be in yet, ready or not. The human story is that we adapted, at real cost, and some of the adaptations — the explicit communication, the written-down knowledge, the cloud resilience — are worth keeping.

I don't know what 2021 holds; a year ago I'd have laughed at anyone who claimed to know what 2020 held. But data work enters it changed — accelerated, distributed, and forced to be more deliberate about the human parts it used to take for granted. Everything moved home this year. Some of it, I think, is going to stay there. And we'll spend a good while yet working out which parts of that we actually want to keep.