3 principles of data journalism you can apply to your Big Data insights
At some of the world’s best-known newspapers, a new style of journalism—data journalism—is generating a growing number of headlines. The data journalist’s core skill? Quickly and efficiently analysing large and complex sets of data, in order to draw out the underlying story.
Put like that, it’s clear that the job of the data journalist is uncannily similar to the role played by analytics experts in industry, as they strive to deliver their own Big Data insights.
And indeed, the data journalist’s tool kit is also very similar to what you’ll find in industry, especially in the realms of Big Data—a blend of open source statistical and data parsing tools, spreadsheets, languages such as R and Python, and a variety of both on-premise and scalable Cloud-based database and analytics engines.
So if you’re struggling to deliver your own Big Data insights, is the answer to hire a data journalist? Well, possibly. But a better—and certainly cheaper—solution might be to simply borrow some techniques from the data journalist’s playbook. Which techniques? And how? Let’s take a look.
Instead of ‘Big Data insight’, think ‘story’
A useful test to apply to a Big Data insight is this: is it actually noteworthy—and can it be summarised in the equivalent of a headline?
Because if it isn’t especially noteworthy, and can’t be summarised in a headline, then—sadly—it isn’t much of an insight. So don’t waste time on it, and press on in the search for the real story.
Such as? Try these:
“40% of our customers never buy from us again.”
“30% of our inventory hasn’t moved in six months.”
“Our most profitable sales are to just 15% of our customers.”
And if you need a fat slide deck and a 30-minute presentation to get a Big Data insight such as one of these across to your audience, then it’s frankly not much of an insight.
If you don’t ask questions, you won’t get answers
But how to find these Big Data insights? Do what journalists—and data journalists—do: ask questions. Sharp, focused, and to-the-point questions.
So don’t wait for answers to leap from the screen. Probe the data hard, with searching questions, applying rigorous ‘so what?’ tests, and maintaining a steadfast refusal to give up.
A good data journalist is a bit like famed TV interviewer Jeremy Paxman, notorious for his refusal to give politicians an easy ride. So if one approach or technique doesn’t deliver the Big Data insight you’re looking for, try another.
It’s the results that matter, not the tools
In this age of low-cost Cloud-based analytics solutions and user-friendly open source tools, there’s a rich variety of Big Data analytics options on offer.
So don’t feel that you have to use one single data analysis tool—SAS, SPSS, R, or whatever. Be flexible, and choose the tool that feels right for the job, and which quickly and efficiently delivers results. Sometimes it might be a spreadsheet, coupled to an add-on such as Big Data-enabling PowerPivot.
Sometimes it might be a tool such as SPSS. And at other times, it might be a business-friendly Cloud-based reporting and analytics tool.
The important thing is to use the right tool. And not the cheapest, or the one that the corporate IT department ‘prefer’ you to use.
Big Data insights: the bottom line
As it becomes ever-simpler and ever-cheaper to exploit datasets of Big Data proportions, more and more businesses will be looking for—and benefitting from—Big Data insights.
But acquiring those Big Data insights isn’t as simple as reading a report, or firing up your desktop PC.
In such circumstances, data journalism has a lot to offer—and we can all learn from cutting-edge data journalists.