Analytics as a core business function: the rise of the data scientist

Organizations have more data at their disposal than ever before, and few would dispute that data analytics can have effective business outcomes. In this context, Amazon is always mentioned as a prime example.

Today, Amazon’s recommendations are based on the user’s wish list, reviewed items and items purchased by other people with similar buying behaviour. This makes the predictive analysis more precise which subsequently results in a positive business impact.

The positive effect is also underlined by the recent McKinsey global survey on ‘The need to lead in data and analytics’. The majority of respondents believe that their analytics activities will have a positive impact on company revenues, margins and organizational efficiency in the coming years.

But actually deriving meaningful insights from the huge amount of data and then taking specific actions is easier said than done. Therefore, more and more companies are pinning their hopes on data scientists.

Just a new buzzword?

Silicon Valley is packed with data scientists nowadays and organizations are competing to recruit professionals with the right profiles, which is already a good indication of what a hot topic data analytics is. But, beyond the buzz, the rise of the data scientist is a fact. So what capabilities do data scientists have that make them so sought-after? “In my opinion, data science goes a bit deeper than data analysis,” responds Andrew Pease, Principal Business Solutions Manager at SAS. “A data scientist uses more advanced techniques to identify business challenges, collect relevant data and publish actionable insights. They’re able to discover trends in data and make meaningful predictions.”

A statistical background helps. However, the really successful data scientist should have a variety of skills. “Data scientists must have an insatiable desire to learn, to innovate and to make things better. Sure, they need to find the data, analyse it, make sense of it and share the results, but if they don’t ask the right questions in the first place, all the data and the best statistical skills in the world won’t help them,” comments Andrew Pease.

According to the McKinsey study, one of the biggest hurdles to an effective analytics program is a lack of communication. Andrew Pease agrees completely: “Data scientists must be able to unlock whatever is inside the data and communicate it to the decision-makers within their organization. They also have to make the complex analytics digestible for the people within the business. Visualization techniques are a great help, for instance, because a picture is worth a thousand words – or lines in an Excel sheet. Making the analytics approachable will give decision-makers a clearer view of what analytics can do for their organization and will hence facilitate the buy-in.”

C-level role

The lack of leadership is another barrier to becoming a truly data-driven organization. A quarter of the McKinsey respondents within high-performing organizations are convinced that ensuring senior management involvement in data activities plays a critical role in the effectiveness of a company’s analytics efforts. “Data science is important for IT and the business alike so the role slips through the net sometimes, in which case it doesn’t receive proper attention from either side. By making data science an organizational, strategic initiative, companies can give data scientists the time and resources they need to be successful.”
Making analytics a core function really is a number-one best practice.

Even though data analytics is top of mind for company leaders, many of them don’t communicate a clear vision throughout their organization. In the McKinsey survey, 38% of CEOs say they are leading their company’s analytics agenda but only 9% of all other C-level executives agree. Those respondents are more likely to appoint chief information officers, chief marketing officers or business-unit heads as leaders of the analytics agenda. “It doesn’t really matter which C-level executive takes the leading role, as long as analytics is a core business function. Up until now, though, analytics has often been perceived as a secondary function in IT departments. The ERP solution is considered to be the backbone, with analytics just something on top. It’s even often seen as a marketing plaything by IT. Making analytics a core function really is a number-one best practice. However, for most organizations, this will not happen overnight. It will take some time to define analytical profiles and they will probably need to demonstrate they can achieve success first before moving into a more senior role.”

In this context, Andrew Pease notes the emergence of the chief analytics officer (CAO): “The CAO has a role on the board and, apart from doing a lot of analytical work, is also constantly evaluating how analytics can play a role in optimizing the business.”

Every organization deals with data analytics in its own way. While some appoint a CAO, others prefer to create an analytical team which may even be cross-functional. “There’s not a one-size-fits-all approach. Organizations have to look at their own specific needs. Some organizations may decide to hire in external analytical competences because they lack the necessary knowledge internally. However, as data analysis is an important strategic part of the business plan, it’s key to internalize the analytical processes at some point,” states Pease.

Actionable analytics

Most larger organizations are already recruiting data scientists. In this era of big data and industry convergence, organizations are realizing that the information contained in the transaction is even more valuable than the transaction itself. “To date, the fastest-growing bank in the UK is actually grocery retailer Tesco. The financial sector is already rife with data scientists, the retail sector is going the same way and over the course of this year there will be an explosion in the demand for data scientists in manufacturing too. With the breakthrough of the Internet of Things, analysing the huge amount of sensor data will be at the forefront,” continues Andrew Pease.

Successful organizational data science is about more than just algorithms. Creativity is essential, not only in terms of how to crunch the numbers but also in the way that data scientists push the resulting insights through to all the organization’s decision points. “If the results are not made accessible and understandable, other people in the organization will find it difficult to act on insights ‘just because the computer says so’.