How to know what your data's not telling you

Share This Page

Data analytics – you know it’s hit the mainstream when you see it popping up in the commercial breaks of your favourite TV shows. I’m talking about Investec’s ‘More than Data’ ad, the centrepiece of a campaign that seeks to show the human side of data.

 

It’s a clever, effective ad that speaks to a real fear we have as consumers – that corporations see us as nothing more than numbers on a spreadsheet. And it relates to a key insight about analytics that many corporations overlook in their rush to jump on the big data train – that data isn’t the be-all-and-end-all of delivering customer experiences, but a tool that depends on how the wielder uses it.

Investec certainly isn’t throwing out the use of analytics and data as a decision-making tool, something that the bank itself admits: “Investec uses sophisticated data analytics as enabling tools. But, we don’t believe in a one-dimensional approach where we place clients into rigid, defined boxes.”

Rather, the ‘More than Data’ ethos speaks to the importance of a human force driving analytics and data strategies, and the understanding that making the right interpretation is as important as gathering the right information. As powerful as data can be, it’s not a mystical crystal ball – the tech equivalent of Paul the Octopus accurately predicting World Cup results. Data can paint us a picture, but we still need to be able to connect the dots.

Don’t book that AI psychologist just yet

To do this, we first need to understand what the current limitations of data are. We know that data can give insights into broad behavioural patterns, such as what times of year make people more likely to be suicidal. This is known as descriptive analytics – condensing large amounts of data into useful insights.

However, predicting how likely an individual is to take their own life is a much hazier affair. Facebook is currently testing an AI that uses pattern recognition to identify at-risk users, as an example of predictive analytics at work.

The next step would be prescriptive analytics, which would recommend the best action, but this kind of model is still in its infancy. Westworld is still a long way away and humans are still far better at making decisions than algorithms. It’s telling that Facebook’s existing suicide prevention tool focuses heavily on giving users the resources to reach out to friends and family who show warning signs of depression. Even if it can identify the people more likely to be depressed, it can’t replace the knowledge and emotional intelligence of a human being in dealing with them.

Understanding the unknown

Alright, so if data is so flawed at engaging with the complexity of the human experience, how can organisations use it effectively without reducing their customers to sets of figures? In all my research of these fancy terms and jargon, the thing that’s stuck with me is the simplest: What you don’t know, you don’t know.

Last year, I had the pleasure of meeting JB Straubel, Tesla’s co-founder and COO. Something that stayed with me was how he described the company’s early days, where they really had no idea where this grand ambition would take them. It’s not like there was any other company out there doing what they were doing – every action was a new step into the unknown.

Despite this, Tesla was clever, using the insights from each new product to bootstrap itself to the next product. When its Model S series manifested an unfortunate tendency to burst into flames, Tesla gathered the data and found that an adjustment in ride height and a battery shield was enough to prevent fire risk. Tesla got to where it was by investing in properly interpreting the data it had – to get to know what it didn’t know.

KLM exemplifies this approach, using data to create unique, memorable experiences for customers. You might remember a few years ago, when they chose to surprise customers with mystery gifts as they waited for their flights. KLM collected data on the lucky recipients through social media, and created personalised surprises for them – a move that greatly contributed to the brand’s reputation as the customer experience guys in air travel. And they did it all years before these kinds of data-driven integrated experiences were commonplace.

 

Jeff Bezos, whose recent data-driven decisions include buying an airline and rocket company, summed it up best: “Anybody who doesn’t change their mind a lot is dramatically underestimating the complexity of the world we live in.”

Figure out where the gaps in your knowledge are – where your human insight needs a helping hand – so you know where you can put your data to work. McKinsey calls it purpose-driven data, where you shape your data strategy around the objectives you’re trying to achieve, not vice versa. And then, when your efforts inevitably go off kilter and you realise that what you don’t know is even more than you thought, adjust your strategy to account for the new information.

If Investec, KLM or Tesla teaches us anything, it’s that it’s time to stop focusing on the ‘data’ part and start focusing on the ‘more than’ part. What are the gaps and how can you overcome them, to deliver unique experiences that make your customers feel like more than just a number? How are you working to understand what you don’t know?

Contact us

* Required field