The goal of the Data Analyst is to problem-solve and drive improvements through learning.  It is rare to have well-established methodologies for the extraction of meaningful insights from data, or clear-cut guidelines for the implementation of novel algorithms in applications such as text classification, recommender systems, and image processing.  Experimentation, trial and error, and iteration is required in order to determine optimal models and their parameters. 

This requires an array of hard skills and technical know-how, as well as soft skills such as being highly collaborative.   Familiarity and appreciation of the scientific method is also key to ensuring that solutions developed meet the requirements of the business.

Companies trying to dip their toes into the Data Analytics space by hiring a single person fresh out of university, or new to the field are finding that this formula does not work.  Recent graduates don’t usually have the business acumen and leadership experience that is required to manage analytics projects in an organisation.  Alternatively, seeking unicorn candidates, with the necessary level of methodical thinking, programming skills, statistical knowledge, and specialised business expertise, is difficult and expensive. 

A more reasonable approach is to define the skillsets your Data Analytics program needs, and build a team of individuals with diverse primary strengths, which complement each other.  Importance should be given to finding and taking advantage of the required skills in a number of individuals, and nurturing this talent in support of business goals.

Once companies recognise the need for a team approach, a common trap that many businesses fall into is trying to optimise teams for productivity gains by functionalising their jobs.  This leads to the definition of specialised roles, such as: big data engineer, statistician, or business intelligence specialist.

This tendency stems from the well-established concept that division of labour drives efficiencies because personnel become highly specialised in their work and are thus able to execute tasks more quickly.  However, in terms of value creation in businesses, the role of Data & Analytics is not to execute.  The Data Analyst does not carry out a predefined set of tasks in support of product or service delivery.  Rather, the goal is to develop profound new business capabilities.

In order to encourage learning and iteration, the focus should be on developing people with broad responsibilities, agnostic to technical function.  An individual joining the company with primarily programming skills must be encouraged to perform diverse functions: from conception to modelling to implementation to measurement.  Likewise, individuals with other primary skills.

The key message is that in order to flourish, Data Analytics teams require first and foremost generalists, with a broad understanding of many numerical and computer techniques, and capable of working across various business functions

The above must not detract from the importance of specialists within the team context.  Depending on the maturity of the Data Analytics program and the associated needs of the business, a move towards more function-based division of labour may be warranted.  However, this will always come at the expense of innovation.

Every cultural and technological revolution throughout history has been characterised by intense periods of innovation and creativity; be it the renaissance, driving rapid evolution in philosophy, literature, and art, or be it the industrial revolutions.  In order to access the full benefits of the data revolution business leaders must build Data Analytics teams, prioritising the development of generalists.  Advantages arise from a generalist team approach.  The cross-pollination of ideas and viewpoints arising from practitioners with different backgrounds, speaking the same language, drives innovation- a necessity for success in a field in a constant state of change.  The most recent developments of which are arising from the advent of Big Data, from the Internet of Things, and from Artificial Intelligence