Munich is a mighty city. Home to not only the Oktoberfest, but also some of the largest German industrial companies, automotive giants and a booming tech scene, it’s hard to accuse it of being behind.
I was attending a startup conference there a few months ago. During one of the discussion panels, a board member of BMW said this:
To tell the truth, we’re fighting for survival. But our challenge is not technological. We have great technical people. Our challenge is cultural. How should we teach our people to work with data?
I was positively impressed with the honesty of that statement. It shows a remarkable awareness of the strength of the transformational trends today. In Munich, at the frontlines of progress, people are concerned.
A few months later, the quote still resonates. It kept me thinking – so how do we teach people to work with data? A lot has been written about the citizen data scientist, democratization of data, self-service BI. There is an overall expectation, that it’s just a matter of time until everyone becomes data driven. After all, we’ve seen this so many times before – so why should it not happen with data?
Why not? Because there are 3 billion computer users in the world, but there are less than 30 million developers. 1 percent.
What I mean by this: I meet managers with a worried look on their face, cautiously confiding they might be missing the data train. They have a growing feeling that there really might be something important hiding in this…
But they don’t know how to work with data. So they can’t even tell if they are indeed missing out.
And I started telling them: Don’t worry about it.
If you didn’t get into data and analytics over the last few years, don’t worry about it.
It’s too late.
If it feels difficult and foreign to drill into reports and explore datasets, if it doesn’t sound like fun to lose yourself in dimensions, metrics and formulas, if creating a single visualisation from a few GB doesn’t feel like a return from a hunt, then don’t worry about it.
If you’re getting into it just because you should, don’t worry about it. It’s too late.
Why? Because the giant gap between data and people will close on its own. Technology will solve it. After helping people learn to work with data for the last few years, I came to belive that
technology will learn to work with people much faster than people will learn to work with data.
To see this, all that’s needed is to look at the trends within trends. “Work with data” often means “looking at reports to find if something’s happening” or “explain why the heck it’s going down”. Working with data is drilling, slicing and generally digging through piles of raw data.
What this reminds me most of is assembly. The low-level coding language that hard-core developers learned in the 90s. (When I was 16 and studying abroad in Florida, my friend Jaroslav Sevcik taught me assembly in a series of mail letters. Paper mail letters.)
Now, how important is assembly today? Technically, incredibly important. Every hardware device needs it. My friend did very well learning it – he’s coding for Google in Munich these days, and it’s likely that his deep, fundamental expertise is one of the reasons. But it’s a layer hidden from 99% of the non-programming population, as well as 99% of programmers. It’s part of every line in that chart above on the “Adoption of Technology,” but it will never be one of the lines.
And that’s how we’ll look at reports, dashboards and data wrangling in a few years. What feels normal today, will become simply amusing. In 1986, this was normal.
It’s still normal today to expect that people need to drill through reports to find meaningful information. It’s normal, but it’s not productive. It’s clearly not something that the majority of people are that good at. It’s just how things are done…for the last 20 or so years.
Same methods, basically same tools, only a lot more data. That’s painful. Like trying to build a skyscraper from bricks.
The current difficulty is caused by the exponential rise of data volumes, and prolonged use of obsolete technology.
That’s also why technology will solve it, and is solving it already. This is a transitory phase. One of the new approaches is what we do at Stories. Our AI automates analytics to provide users not with data, but with the most important insights in their area. Only 3-5 stories per day. Not data, not reports, but information. Stories like this:
They don’t require specialized data knowledge to work with. The data analysis part is automated. It reduces information load by a factor of 1000. The business user can focus on what they know best – doing business. Making decisions. Taking informed action. Learning and improving.
That’s why I say – if you are not the analytical type, don’t worry about picking it up in 2017. Don’t be intimidated by it. It’s a complex job now, but it also isn’t necessarily one that’s done best by people. The tech companies are already on it. Focus on the business, the action, the customers. And simplify.