The emergence of technologies and methods comprising a big data infrastructure and its applications has positively influenced business enterprises and society as a whole in economic, political, and social dimensions. Every single day new projects and initiatives are successfully carried out by enthusiastic professionals who bring their creativity, intelligence, tenacity, and knowledge to the increasingly promising area of data science and its related disciplines.
The success and growing interest in big data and analytics has caused an extreme imbalance between supply and demand for skilled professionals in workplaces. Moreover, according to a well-known study published by the McKinsey Global Institute (MGI), data science professionals are not the only scarce human resources to work on big data projects. According to this MGI research, many organizations are prone to missing opportunities and competitive edge simply because their corporate executives do not possess the necessary vision and skills to understand the power of methods and technologies, such as machine learning algorithms applied to a vast and diverse amount of data that would greatly improve business performance.
Building a Strong Analytics Corporate Culture
Despite numerous recent programs from universities, colleges, and other institutions to provide education on data science from functional and technical perspectives, most organizations still struggle to hire professionals who meet the essential set of requirements that would lead to a superior and effective performance on data analysis tasks. The shortage of professionals who meet, partially or completely, the functional prerequisites imposes additional challenges for corporate executives and project managers. Nevertheless, the lack of maturity in the use of big data applications and data science in general has also led to frustration in the recruitment processes of companies looking for data analysts and data scientists. Very often, poor performance of newly hired professionals is an inevitable consequence of unrealistic expectations and/or an incomplete or wrong profile definition for the role assigned to them in the organization.
In order to avoid common mistakes and pitfalls in analytics initiatives, especially in human resources management, it is highly recommended that companies, during their early stages of the implementation of enterprise or departmental solutions, strive to get the benefits of consulting, support, and training from experienced analytics business services providers, while building their own expertise and corporate culture in this area. The support of highly skilled data science consultants may be extremely useful, not only to help define the right profile of the data analyst role, but also to recognize people who are already part of the company’s workforce and may move to new positions to better contribute to the deployment of analytical solutions. In addition, experienced professionals may also be handy in identifying key areas where big data applications would have more significant, immediate, and positive impact to the enterprise.
One common mistake that has led many companies to frustration when they are looking for analytics professionals is the misconception that it is possible to get all knowledge and skills assigned to a data scientist role from one single candidate. As demands and the nature of big data applications become highly diverse, the requirements for such high-skilled workers to effectively perform these functions may also vary significantly.
A Framework to Outline the Data Scientist Role
In fact, the functions of a data scientist extend across a dimension that connects two extremes: data sources and business areas. In the first extreme of what may be called a data science application spectrum, the most important technical knowledge is the ability to deploy, manage, extract, and integrate information from various heterogeneous data sources in a way that they could be enriched, modeled, queried, and visualized by users seeking answers to relevant questions in their decision making processes. At the other extreme of this spectrum, there are professionals who directly support corporate or local business executives to get insights from data sets previously prepared and readily available to them, usually exploring visualization and simulation tools. They should have an in-depth knowledge of a specific business domain, the role they play within the organization, and how they can help in delivering good results for the enterprise or one particular group or department.
In the midst of these extremes, a data analyst is also responsible for defining and customizing mathematical and statistical models designed to solve specific business problems and improve the overall performance of business processes throughout the company. This could certainly be a very technical endeavor. However, thanks to a new generation of data discovery tools and software applications, even business analysts have been able to use sophisticated machine learning algorithms in data mining processes that would not have been possible without extensive computer programming and statistical modeling skills just a few years ago. These tools may also implement visualization functionality, which allows users to easily create an abstraction layer for building, testing, and fine-tuning these models in specific scenarios. Still in the middle of these extremes, analytical solutions may be designed, developed, and made available for general usage in the form of (cloud or on-premise based) software systems, which requires additional expertise in application development and user interface design.
Regardless of the maturity level of an organization on the usage and implementation of analytical solutions, enterprises should face reality and go after analytics professionals to carry out the job assignments and duties briefly introduced above. However, as stated previously, it is very unlikely that one single person would portray in his/her resume all of the skills and expertise necessary to do them effectively. Therefore, a more reasonable approach would be to build a team of analytics professionals based on a strong academic background and solid work experience in all data science aspects, which involve areas such as data management, mining, and visualization, as well as application development, deployment, and lifecycle management in some cases. All technical background should be complemented with the domain knowledge necessary to build models, and this combination would be an important part of the collective skills gathered in the team building process.
Human Resources as Part of Overall Analytics Strategy
Unfortunately, building a team with analytics expertise to deliver sound results and respond quickly and efficiently to an ever-increasing demand for more insights from data is not the only challenge for C-level executives in a big data environment. Very often, business leaders also face dilemmas to find the best organization structure to deliver the promised results and insights of analytics. The most common issues may involve decisions such as (1) centralized or decentralized approach for managing and carrying out analytics projects, (2) single technology platform or solution provider or best-of-breed solution for each type of problem or business area, and (3) implementing a business intelligence competency center (BICC) led by a Chief Data Officer (CDO) or having the analytics team reporting directly to the CIO or CFO. Discussing which structure best fits the particular requirements of one’s organization would be a good topic for the next article.
Without question, having the right talent on board is essential to carrying out any business initiative, especially those which involve high levels of innovation on both technology and business processes. Nevertheless, successful implementation of big data infrastructure and methods requires an analytics strategy and organizational structure that consistently and proactively address business objectives and stakeholders’ interests. This strategy will set the guidelines for businesses to become more competitive by taking advantage of the extraordinary opportunities enabled by data science technologies. Hiring and retaining the best people is certainly a crucial step in helping organizations overcome further challenges and get the best of out of their investment in analytics.
