Big data is often considered to be a high-tech endeavour, something that requires huge computing power. It needs to have significant technology just to get off the ground and requires huge financial investment to get even the most rudimental understanding of new capabilities made possible through effective analysis.
However, this is not true. Big data can be leveraged for almost nothing, but you need to have the right team in order to do so.
The problem for almost every data-driven company in the world today is that these people are hard to find, and the demand for their work outstrips the supply. McKinsey predicts that in the US alone there are 140,000 to 190,000 analyst positions that remain unfilled. The problem with this number is that often the people who are tasked with filling these roles have little specific skills in the area. This isn't fair on the company or the individual, but is increasingly being seen across the workforce.
Little is being done to address the shortcomings in the market though, and TechCrunch claims that of the top 100 universities in the world, only 29 have data science courses and only 6 of these offer a course at undergraduate level. With the average class size amongst these being just 23, it is easy to see why the skills gap exists.
With this restricted workforce it is vital that you make the right hiring decisions when bringing somebody new into the team, to not only make sure you are getting the best, but also that they are not going to need replacing within 6 months. The global shortage means that bringing people in from other areas to fulfil roles in a team is common, and when done properly can create an innovative data science team, such as those found at Uber and Airbnb.
So what kind of skills that already exist in the company can be utilized?
Being able to pick up potential correlations and patterns is essential in any data work, and luckily this is something that should exist in many areas of the company. Given that this kind of work is going to be dealing with numbers, rather than specific business problems, there are going to be some existing roles that are more suited to here, such as…
Existing departments: Accountancy, Analysts.
The ability to manipulate coding and algorithms in order to change how data is analyzed is key to an effective data science programme. Given that this is a predominantly technical role, there should be technical expertise to do this, even at the most elementary level.
Existing departments: Web Development, App Development.
Another key element of being able to effectively manage a data science team is making sure that the data being used is fit for purpose. This means regular auditing and testing, and is something that is done to varying degrees elsewhere in the company; this same skill set could therefore be utilized in a data science team.
Existing departments: Accountancy, Marketing.
A key part of any data science programme is naturally the ability to communicate any findings that are uncovered with the relevant people. This means being able to adapt language towards others within the company, whilst making sure that the findings are actionable and clearly understood.
Existing departments: Marketing, Copywriting.