Data Science in the Corporate World.

Organisations find themselves at the dawn of a new era of unruly change. With access to new technologies, as well as unprecedented levels of information, businesses are transforming at an extraordinary rate. 

But in that respect, they aren’t alone. Thanks to data science and machine learning, virtually every corner of society has been reshaped in some way or another. Whether it’s via the internet, or our mobile phones, we have access to huge data sets which offer us an abundance of potential insights into how people and ‘things’ interact. Looking at businesses and organisations specifically, this explosion of information has enabled countless firms to implement automation, and act upon insights and intelligence at a scale that was previously unattainable. 

And, as data science and artificial intelligence continue to supply businesses with new technologies capable of improving productivity on non-physical work, among other things, the next generation of technology-driven change to organisations and work is likely to be wide-reaching and profound.

But with the backdrop of last year and a half in mind, the need – and the desire – for organisations to harness newfound insights and reimagine certain functions has arguably never been more widely felt. Making the most of these new technologies will therefore be critical to future organisational success. But this will take a new kind of thinking, as systems start to affect not just lower paying jobs, but also better paying, higher-skilled jobs that require significant training. 

While on the whole, data science and business analytics have become reliable, recognised tools among those organisations that are at the cutting edge of their field, for many, the idea of data-driven artificial intelligence (AI) and machine learning technologies still sounds overwhelming. But that doesn’t take away from the fact that, if applied effectively, data science can be used to not only rethink the work companies are doing, but also the boundaries of firms themselves and the architectures of industries. 

While the list goes on, here are three core takeaways for those looking to harness data science in a business context. 

Technical knowledge is a must-have 

It goes without saying that, to use data science effectively in any endeavour – business or otherwise – it starts with technical knowledge. That may seem like a statement of the obvious, but there remains a lot of confusion in the marketplace as to what skills are needed to call someone a ‘data scientist’. With the hiring of professionals in this field having spiked as of late, businesses need to get better at knowing what they should be looking for. And, importantly, aspiring data scientists need to know their stuff. That’s where institutions like Imperial College Business School come in.  

At Imperial, the Data Spark scheme equips students, studying across a number of subjects such as the MSc Business Analytics programme, with real-world skills by engaging them in real-word problems. Working in teams with an academic mentor, as well as a client sponsor, students look to address a particular issue and deliver their results directly to the organisation with the problem. And, having completed projects across a number of industries, such as the not-for-profit sector, financial services, healthcare, energy, aerospace, and professional services such as consulting, law and accounting, students, upon graduating and starting their careers, are able to harness tried-and-tested methodologies and bring legitimate experience to their roles early-on. 

And its real-world exposure like this, under the supervision of experienced academics and practitioners, which gives professionals working in the field of data science the skills and understanding to effectively carry out key tasks, such as supervised and unsupervised machine learning, as well as the know-how for dealing with uncleaned data, which leads me on to my next point…

Working with dirty data is often unavoidable

While in a classroom or laboratory environment you may be afforded the luxury of having data that’s been put together well, in the real world data sets come with problems. Using new techniques of ‘data learning’, it is possible put even problematic data to work safely and economically. But it still requires you to demonstrate what we at Imperial call ‘data entrepreneurship’ – finding ways to get around challenges – such as uncleaned data – by being resourceful and creative. If you can’t do this, you won’t get far as a data scientist. 

Nonetheless, the sobering reality of uncleaned data reminds us that we can’t trust AI-based systems if we don’t know about the data and model behind it. It’s therefore crucial that, as a data scientist, when dealing with data-driven AI and machine learning technologies, you ask those important questions: Is the data ethically sourced? Is Are systems replicating human behaviour that needs to be challenged and changed? Is the model achieving required accuracy or performance? Is the model validated for conditions we face now? 

In asking those questions, you open yourself up to doing more thoughtful work and, in doing so, distinguish yourself from the unscrupulous modellers out there. 

Know more than just the data science 

While technical skills and understanding related to data science are fundamental to utilising AI and machine learning effectively in a business context, you also need to have an understanding of – and an interest in – the broader industry you’re operating in. 

Whether its retail, financial services, or even healthcare (the list really does go on), you need to understand the nuances of your sector – and your organisation. And if you don’t have that interest, or that understanding, then you need to have someone who does. To work in data science teams, you have to be at least ‘domain-curious’. 

Now, that’s not to say that you need to be an expert on the industry you’re operating in. But the professionals that have the greatest impact are the individuals who aren’t simply waiting for a problem to arise, or for an executive beyond the data science team to flag an issue. They’re learning enough about the business to bring their own ideas to the table, identifying problems to solve that others, who don’t have a data science background, aren’t thinking of. 

Therein lies the value of data science in business a context: always working to find new problems to solve, challenges to overcome and insights to uncover in the pursuit of making your organisation more effective, efficient and resilient in a world that’s always changing. 

Dr Kennedy is Associate Professor of Strategy & Organisation at Imperial College Business School, Director of Imperial Business Analytics, a research lab in Imperial College London’s Data Science Institute (DSI), and also a deputy director of the DSI.