The data scientist has become one of the most celebrated employees of the Internet era. But Duncan Keene, UK managing director of ContentSquare, explains that an evolution in their job role is sorely needed.
The importance of big data to global business cannot be underestimated. As the potential for the proper application and interpretation of data becomes clear, the role of the data scientist has been called ‘the sexiest job of the 21st century’.
But those on the lookout for a data specialist find themselves in a fierce war for talent. Research by McKinsey has shown that there is a global shortage of data scientists, with the US alone facing a shortfall of 140,000 – 190,000 ‘people with deep analytical skills’, in addition to an astonishing 1.5 million data analysts and managers. This seemingly chronic shortage makes the work that data scientists do for their employers all the more significant.
The issue is, even if a company manages to hire a data scientist, many discover that they don’t know how to best use their considerable skills. A recent MIT Sloan Management Review article argues that even those businesses with established data specialists ‘struggle to realise the full organisational and financial benefits from investing in data analytics’.
So what’s going wrong?
For a start, the data scientist role is often created by a board that doesn’t fully understand the potential of big data. Despite their highly specialised skills, if data scientists aren’t quickly integrated into the day-to-day running of the business they can become isolated, unable to communicate the value of their insights, and unable to help inform company strategy. This isn’t the fault of the data scientists themselves, it’s a deficiency within the structure of a business not fully prepared for a full time data scientist role.
Online retailers are a prime example of the transformative effect of data science. Companies like Asos, Zalando and boohoo.com have been able to swiftly achieve significant sales volumes through an in-depth understanding of consumer behaviour.
Ecommerce data pioneers have created increasingly sophisticated analytics, measuring everything from the user’s cursor movements and hover times to patterns of behaviour and ‘broken journeys’, when the shopper fills the basket but doesn’t purchase.
Yet even with huge volumes of product performance data, and a data scientist to interpret it, retailers would admit they still struggle to generate maximum value from their data. McKinsey, the consulting firm, has found that retailers using big data to the fullest could increase operating margins by more than 60 percent. For some retailers it is clear that ongoing data reporting has become too routine, with findings being at worst ignored and at best only actioned infrequently.
For online stores, quantitative data such as sales volume or website bounce rate are useful statistics for measuring product performance, but such metrics struggle to reveal why an item is actually selling. The issue here is that data scientists are being asked to perform a task that shouldn’t be part of their job. Data scientists are there to analyse complex information, not to act as gatekeepers between basic company data and the rest of the business. This data should be easily and readily available to all.
What’s more, data scientists can find themselves struggling to communicate their worth in complex organisations. Because statistical data analysis is a specialised and challenging field, it can be difficult to get ‘buy in’ from the majority of the workforce where this specialist knowledge is not present.
The solution to these issues, as counterintuitive as it may seem, is to strip back the role of the data scientist. The role of gatekeeper between data insights and the wider organisation can and should be delegated to an automated tool.
There are many advanced visualisation tools which help make basic data more understandable. The key is to use software to translate quantitative data into qualitative insights that do not require specialist qualifications to interpret and will therefore deliver clearly actionable data insights to a far wider audience.
Tools such as heatmaps and interactive charts can go a long way to increasing employees’ general awareness and understanding of data. The overall objective should be to move data about broken customer journeys closer to those who actually have the knowledge and ability to action that information, the product managers.
This would be good news not just for product managers, who would be able to apply data insights to product development in a much more precise way, it would also be of benefit for business leaders, who can make a more precise decision on marketing spend, and the data scientist knows that his/her work is adding value to the business.
We’re seeing a real cultural shift in the way businesses approach their data and therefore the role of the data scientist needs to evolve. Business leaders need to re-focus data scientists’ formidable skills to advising senior management on business strategy and away from mundane, everyday data tasks.
There are powerful analytics toolsets available on the market which can automate many of the more simple tasks currently being directed at data scientists, and these tools have a potential to fundamentally shift business attitudes to data. If the business ‘rank and file’ can be given more ready and intuitive access to the business’ underlying data, better and more informed decisions will follow.
By Duncan Keene
UK managing director