A lot has been written about data literacy in recent years as organizations seek to hit that sweet spot of having game changing technology and people with the skills to use it to uncover insights. It’s all about becoming a truly data-driven organization. But I believe the level of data literacy your organization needs to do that really depends on what your employees and individual teams are trying to do. And I would argue everyone is data literate to some degree.
What is being data literate?
Let’s start with a definition. For me, at its simplest, data literacy means that people have the skills to interpret insights which helps them to do their jobs well.
For a data scientist that might mean a lot of analytical thinking and specialist knowledge. But on the other hand, a shop cashier would have a completely different level of needs.
That’s the heart of the issue really. No organization on the planet is going to train everyone it employs to have the skills of a data scientist. And why would they – when all you need is enough for them to do their jobs well.
“No organization on the planet is going to train everyone it employs to have the skills of a data scientist.”
Understand how different teams use data
You need to think what data you have which enables your teams to do their job – and you really don’t need to go above and beyond that. A lack of data literacy is often the crutch of the BI analyst industry to excuse poor technology that doesn’t actually help the business user. And we are all guilty of it.
For example, over the years I’ve had some amazing ideas about data which gets data scientists all in a froth. But if you put it in front of a business user they’re like – meh. I don’t think so. And 9 times out of 10 that’s because they can’t relate it to their day to day job.
Design for user needs, not data scientists
It’s important to design something which actually helps the business user and switch your thinking away from blaming them for not being data literate enough.
That’s not redefining data literacy though, as the shop cashier we mentioned with the cash register still needs to be able to read the numbers and know the data. You have to be data literate to do that.
Likewise, if you’re running a marketing department you need to know where your leads are coming from. You need to know about attribution and how your data is collected – and crucially you need to know about the pipeline.
So ultimately, they are very different set of skills. But in both cases, you don’t need to be a data scientist. You are matching the skills and technology to the needs.
Individual learning styles are important
This is something a lot of data scientists overlook. They assume everyone wants to consume data how they do – numerically.
But you can learn in lots of different ways. For simplicity’s sake, let’s say that’s learning numerically, by reading, or by talking.
Think of the data and the presentation which will be the most useful and easy for the users to work with. For example, lawyers will always be narrative driven as they live and breath stories.
This all comes back to where most people actually work. Most people work within a process, and application developers build tools to make those processes easier. And the ‘doers’ live in those product.
For example, sales people spend their whole day in Salesforce. They don’t go anywhere else. And as a trend we’re seeing people injecting data analytics into these processes.
So, where does this definition of data literacy leave us?
I think the reality is you will always need those super smart data people. They won’t go away. But understand who those tools are for. And think of different tools for different people. It’s all about guiding people to reach their potential and use what they already instinctively know.
To me that’s where I see the answer. It’s using data to drive behavior – not hoping to change people’s behavior to use data.
About the author
Glen Rabie is CEO and co-founder of Yellowfin, an enterprise analytics suite that combines industry-leading automated analysis, storytelling and collaboration.