Applications iStock 000006482851Small

It can still raise a question over what it’s about, but ‘data scientist’ has joined the vocabulary of the business world and now promises to influence the skills agenda for the UK industry. The rising demand to exploit rising data levels is prompting organisations like e-skills UK to look at how occupational standards can be developed to fit the bill, and employers have told the body that they want universities to provide more relevant degree courses.

The potential of the data scientist has become more pressing as more organisations have appreciated what they can do with big data and look for people with an understanding of algorithms; but there is a debate over the difference, if any, from a data analyst.  

“We find a very great blurring of the job roles, titles and skills,” says Tony Venus, head of standards and qualifications at e-skills UK. “If we look at different titles in different organisations, quite often 80-90% of it can be common between the two.”

But he says that, while the analyst needs to understand data structures, the data scientist needs a good understanding of the business context and the potential value of the data.

“A data scientist would have some business understanding, not full strategic responsibility but awareness of what the data can add in terms of business value.”

A similar perspective comes from Nick Millman, managing director big data and analytics for consultants Accenture. “The data scientist will use predictive modelling and data analytics tools to find insights in the data sets,” he says. “One of the key success criteria is helping the business understand it to take action.”

He adds: “Data scientists needs to be good at explaining the outcomes of the data science to the business so it becomes actionable.

The benefits to the employer are largely in increasing the predictive capabilities of the data. Examples include reducing risk in a credit portfolio or taking action to reduce customer churn, and it becomes even more valuable when used to sharpen real time reactions.

Millman cites an example from the telco industry of a data scientist building models to segment customers, partly by demographics, their billing and usage records for the service and data from other sources, then working out predictive models on which are most likely to give up subscriptions. It could then provide incentives to renew that are relevant to individual customers or maybe, if they are not profitable, decide to let them go.

Other examples are in the real time analysis of social media sentiment on an issue, or about a company, reacting to retail trends in something close to near time, and spotting anything suspicious on financial markets in immediately rather than retrospectively. Data scientists can also support long term planning, such as reducing risk in a credit portfolio.

Mike Cook, chief information officer of address management software company Postcode Anywhere, says: “Data scientists are needed to help calculate the best possible timing for interactions with potential or existing customers, and what gives them the ‘Wow! moments’ when interacting with a company. This is important to ensure customers feel you’re taking the time to personalise their experience with the company.”

He also sees benefits in implementing systems that give more people within an organisation access to key data that can provide value. For example, it can give operational managers access to real time data on staff performance to help make efficiency improvements.

The key feature in all of this is to provide actionable insights, making an organisation more responsive to changes in its environment and giving it a competitive edge. There may still be questions over when to use the term data scientist, but the demand for the relevant skills is going to increase.