Solving the people problem: Key part of getting the most out of big data

By Rich Clayton
Apparently, less than one in five managers actively use analytic tools.   Instead, based on historic processes, they rely on IT to build reports for them, from legacy systems that could be described as slow and stoic. In today’s dynamic, data-driven world, this approach will hardly set the world on fire.  
Why is this still the case? A joint investigation by Teradata, Forbes Insight and McKinsey & Company found that for every dollar invested, there’s a 10x return on analytics initiatives. So where’s the problem?
Increasingly, any reluctance to adopt analytics is less about the technology available.  The advent of powerful, cloud-based, data visualisation tools is putting insight into the hands of everyone – not just the highly skilled data scientist.  As a result, curious, non-technical business consumers can start exploring data without any knowledge at all – and, where this is the case -- amazing things are starting to happen.   
So if IT is a slower option and the technical barrier to entry is low, then why aren’t more managers using analytics?

Resourcing data analytics
Today, with the advent of big data, organisations are looking beyond IT and increasingly are turning to data scientists for insight. There are a number of challenges with this.  Data scientists are in short supply. Management consulting firm, McKinsey & Company quantifies the extent of how resourcing is holding backing the potential of data analytics. 
By 2018, there will be a shortage of 1.5 million data savvy managers and a 150,000 gap in available data scientists in the US alone.
Due to the laws of supply and demand, these highly skilled and educated individuals tend to be an expensive commodity; one that is hard to sustain. Only the largest organisations have the funding and workflow to keep a team of data scientists busy. 

The data savvy business manager
Data-savvy managers inside the business who can understand the business and leverage analytics may be what is needed.  
For example, in sales and marketing, it’s about seeing the demand signals, understanding campaign performance and customer experience. In research and development, it’s about seeing a return on investment across the portfolio of projects. In Human Resources, it’s about seeing patterns in turnover and talent progression. In Finance, it’s about visualizing profit and growth drivers.

Changing education
For me, a large part of the issue centres on education.  Few tertiary institutions around the world have grasped the changing, real-world need. In my experience, when you ask where in the institution data science is taught, almost invariably it is within statistics department.  Universities are trying to remash their computer science and statistics courses in an attempt to catch up.  While a laudable activity, what we really need is to entirely reform the analytics curriculum to bring in business skills and prepare young people for the new workplace. 
But let’s not lay all the blame at the foot of the universities.  Business professionals are ripe for expanded and ongoing education of data analytics.  
We need people who can cross domains, and bring longitudinal and latitudinal perspectives to data and business decision making.  Great leaders marry communication with analysis, and the resulting story leads to great outcomes.  The inverse is also true: if you can’t tell the story nothing happens, and this seems to often be the case with big data.  We see many successful pilot projects taking place, but few move to production.
 We need to join the soft skills with the data science.  Fortunately, there is a whole wealth of training courses out there on big data, with many available online and for free from the likes of Coursera.
And attacking the problem at these different levels will have a number of great affects.  

Adding diversity to counter bias
Typically, older managers have deep business experience and company knowledge have got to where they are based on gut feel – partly by the fact that they haven’t had the wealth of insight now available at their fingertips.  However, in today’s world, we are asking for the opposite to be true and for them to trust the data.  
In addition, often innovation rests with people who can come in with an outside view – “Why can’t we do this?” Whereas, an expert might say “We can’t for [whatever] reason”.  So as you can see, bias, however well intentioned, can hold executives back from taking the next step. In fact, intuition stops analytics.
At the other end of the scale, the millennial workforce, who by virtue of age, can lack the industry experience and company knowledge is well tuned to modern technology.  Young professionals bring new ways and ideas on how to work and are challenging existing analytic processes and systems.  Generally, they are more able to whip up an analysis quickly, and typically expect all to follow.
In the new era, where we’re learning to trust the data, what’s critical is a workforce with a diversity of not just experience, but also skills. 
As an extreme example, the University of Pittsburgh is hiring people with music qualifications to avoid expert bias in its study of breast cancer – because you can’t study the human genome sequentially.  

In summary
We’ve got to get smarter about our approaches and our development of the skillsets needed to capitalise on data analytics. Most of the data we’re starting to extract now has not been looked at before. As the data becomes more complex, our ability to derive value from it becomes clouded to the extent that we don’t even know how to ask the right questions.
A multigenerational, multiskilled workforce should lead to more interactions around data within, and with greater benefit.  That is not to say there is no requirement for data scientists but I predict that future analytics leaders will have their education more based in liberal arts training than solely pure maths. 
(The writer is VP Business Analytics Product Group at Oracle)