We're not all data scientist - and never will be


By Richard Badley

Picture this: Rows of people sitting silently in front of a mass of computer screens. Staring blankly at millions of lines of data passing in front of their eyes. Searching for anomalies, trends and ever hopeful to understand more, desperate to identify opportunities and maximise efficiency. A world like the one revealed to Neo in the Matrix when he realises he is ‘the one’. Beautiful binary bliss.

I’m not describing a billion-pound trading desk, but an imaginary place that, according to all the hype around data, will exist for everyone. A world where if you don’t live and breath measures and metrics you’re guaranteeing failure for your business, and personally committing career suicide. We’re led to believe that if we don’t operate in absolute real time we’ll opportunities, or adversely, haemorrhage the limited budgets we have through incredibly misjudged decisions.

Will this really happen? No. Of course not. Will we ever live in a world like this? No.
Why? Because we aren’t all data scientists, and we never will be. People are intrinsically different and we simply won’t all do the same job, in the same way. The assumption of us all being fixated with figures also doesn’t allow for any application of experience, preconceptions and the power of simple ‘good judgement’.

So why all the fuss?

Data isn’t new but it is now more prevalent than it’s ever been. Traditionally controlled by specialists in dark corners of organisations, data has now broken the barriers and filtered across entire organisations. This exponential growth has come from higher volumes of data, a wider variety, and it all coming at us at a faster velocity.

This leaves us in a tricky situation. We aren’t all going to be data scientists, but we’re surrounded by more and more information. So we need to do something about it. We need to simplify.

Ignore the scare mongering and hype, step back from the precipice of ignorance-forced failure and focus on what we really want to understand, achieve or improve. Look at your objectives and goals, and not the data around you. Because, let’s face it, the questions and challenges that we encounter daily are generally the same as they’ve always been. Therefore, we know what we want. What is different now is that there’s a strong chance the answers may lie in data.

There are some basic rules of data engagement:

1. Always start with a clear objective.

2. Use common sense (you’re smarter than your data).

3. Don’t answer the wrong question – it’ll end in disappointment.

4. Don’t be scared to ‘spend time’ with your data.

5. Keep it simple, because simple is beautiful.

6. Ensure the outputs make sense. However complex the process might’ve been to get you there, it doesn’t mean people want to see it. A pie chart is fine, even if it took months of complex work to get there – if it answers point number 1, you’ve done it.

When data analysis works, it’s exceptionally powerful. Take in-house recruitment for example. Through intelligent use of data, it can significantly increase quality-of-hire, decrease both cost-per-hire and time-to-hire, and even make it possible to track channel performance in real time.

But we do have to make a choice. Either we sit and wait for others to take the reigns. Or we have a good go at getting some value from the assets, information and insights that sit behind the screens.

Let’s face it, of the entire UK HEI cohort only 2.2% are studying mathematical related subjects (HESA). Now, I’m not data scientist (and I also appreciate that people working as data specialists can come from a number of backgrounds), but these figures suggest we’re going to have a skills shortage. Therefore, perhaps we do all need to have a little go at it.