Who are today's analytics professionals?
How do we frame the skills and professional context of all analytics professionals? How do we articulate the value one brings to multidisciplinary conference calls? And how do we orchestrate the strengths and challenges of team members to ultimately serve clients?
The answers to these questions are essential for delivering small or large analytics projects on target, on time and on budget. Nine fields of study exist for corresponding professionals who contribute significantly to analytics projects, their key value propositions and the challenges and high-level suggestions for categorizing and organizing the rapidly growing analytics resource pool.
What exactly is the field of analytics? Analytics is the methodical, computational analysis of data or statistics. It involves finding meaningful patterns in data and the ability to communicate them. In addition, analytics constitutes a multidisciplinary field that incorporates computer programming, operations research and statistics to quantify performance.
Massachusetts Institute of Technology (MIT) Center for Digital Business Fellow, Dr. Thomas Davenport, believes “the field of business analytics was born in the mid-1950s,” and we are currently in the era of Analytics 3.0. (For a more detailed snapshot of its history, see the graphical timeline, “A Brief History of Data Science and Evolution of Big Data Technologies,” created by Mamatha Upadhyaya, Analytics CoE Lead for Capgemini India.) Whether one is new to analytics or grew up in one of the disciplines, understanding the nine fields of study, how they overlap and how they are dependent on one another to solve complex business problems is quite important.
Jean Francois Puget, an optimization mathematical guru at IBM, thinks “the value of analytics comes from the business decisions it enables” and automates. Because analysts should always start with the business questions first, and not the data, veteran business analysts and domain experts need to be engaged from the beginning to the end of a project. That engagement includes the stages to discover, develop, champion and capture the value that analytics can provide. Depending on the business problem at hand, business analysts may be domain subject matter experts (SMEs). For example, they may be marketing, operations, supply chain or workforce SMEs, industry SMEs such as in retail, energy and banking, or both such as a supply chain expert in Canadian liquid pipelines.
Key business analyst value propositions include business value maps, context, vital pain points, key performance indicators (KPIs), model requirements, process and IT re-engineering, testing and calibration, and so on. And key challenges associated with this discipline include the struggle in documenting tribal knowledge and translating business requirements into mathematical formulas, agreeing on trade-offs among competing business priorities, calibrating output of analytical models and sometimes maintaining an all-or-nothing attitude.
In this area of study, strategic and financial skills—typically demonstrated by MBAs, chartered financial analysts (CFAs), chartered accountants (CAs) and other financial professionals—are very important to create a business case at the front end. They also are important for conducting periodical project progress checkpoints and ensuring executive alignment and complete business value quantification at the back end of an analytics initiative. For example, Procter & Gamble decided to collect data from its manufacturing facilities in the 1990s to understand how products and machines could be optimized. Specifically, business value quantification was at the top of the organization’s agenda to “create operational efficiencies that would contribute to healthy” earnings before interest, taxes, depreciation and amortization (EBITDA) margins.
Key value propositions for practitioners of this discipline include business cases, project funding approval and compliance, value based go and no-go decisions or changes in direction, benefit quantification and more. Unfamiliarity with optimization, simulation or predictive model development; testing; or implementation of complex IT projects are among key challenges within this discipline.
According to The Institute for Operations Research and the Management Sciences (INFORMS), the field of operations research overlaps with other disciplines, notably industrial engineering and operations management. The essential contribution of this discipline is the construction of mathematical models that attempt to describe the system. Often, these models are concerned with determining a maximum—profit, performance or yield—or minimum—loss, risk or cost—of business objectives.
INFORMS is the largest society in the world for professionals in the field of operations research, management science and analytics. It governs the Certified Analytics Professional (CAP) designations, and it hosts an annual conference on business analytics and operations research.
Problem-solving techniques and methods applied in the pursuit of improved decision making and efficiency—such as simulation, mathematical optimization, queuing theory, Markov decision processes, neural networks and decision analysis—characterize key value propositions within this discipline. Challenges within this field of study often translate to not getting involved with politics along with organization change management, project timelines and budgets, and core IT-related issues such as user interfaces (UIs), databases, integration and infrastructure.
Particularly with regard to inferential analysis of population characteristics from sampling, statistics encompasses the math applied to collecting, interpreting and organizing numerical data. Statisticians apply statistical thinking and methods to a wide variety of scientific, social and business endeavors such as astronomy, biology, economics, education, engineering, genetics, marketing, medicine, psychology, public health and sports among many others. The American Statistical Association (ASA) is the world’s largest community of statisticians.
The key value propositions that statisticians bring to the table are descriptive statistics such as mean, standard deviation, frequency and percentage and inferential statistics—such as answering yes or no questions about the data, or hypothesis testing. Others include estimating numerical characteristics of the data such as estimation, describing associations within the data such as correlation, and modeling relationships within the data such as regression analysis. Inference can extend to forecasting, prediction and estimation of unobserved values, as well as the extrapolation and interpolation of time series or spatial data and data mining. Challenges within this discipline are very similar to those for operations research professionals.
Data scientists advise executives and product managers on the implications of the data for products, processes and decisions by enabling the shift from ad hoc analysis to ongoing conversation with data. This area is a hybrid practice of research and execution along with the application of the start-with-data-first philosophy. “As they make discoveries,” Davenport says in his Harvard Business Review article, “Data Scientist: The Sexiest Job of the 21st Century,” “they communicate what they have learned and suggest its implications for new business directions.” And those directions may be new revenue models, cost control or regulatory compliance.
The industry is still defining the data scientist role, and data scientists are creating new deliverables and tools every day. As Davenport mentions, Yahoo was among those firms that employed data scientists early on, and it was instrumental in developing the Apache Hadoop framework. Similarly, Facebook’s data team created the Apache Hive language used for programming Hadoop projects. Data scientists face challenges very similar to operational research professionals and statisticians, but they tend to be more core-IT savvy.
Visualizing information such as data, facts, ideas, issues, questions, statistics and subjects, and all while using words minimally, is a rapidly growing and essential discipline in its own right. Data journalists and information designers such as David McCandless, author of Information Is Beautiful, refer to visualization information as combining “language of the mind” with “language of the eyes.” His data visualization process includes the conception and the generation of good, interesting ideas; researching and curating juicy data; executing and selecting appropriate and effective visualizations; and designing and beautifying impactful charts and diagrams.
Key value propositions in this area include creativity in displaying information visually and making the patterns that are found clear and compelling—for example, information design, hypothesis, charts and diagrams. The challenges are the same as for data scientists.
Computer science—or information management
A computer scientist specializes in the theory of computation and the design of computational systems. Computer science is the study of the access to and acquisition, communication, processing, representation and storage of information that includes the expression, feasibility, mechanization and structure of its underlying methodical procedures. Under the information management umbrella are many sub-disciplines such as business intelligence (BI), data acquisition, data integration, data marts and warehouses, Enterprise Content Management (ECM), streaming computing, transformation and performance management.
This field of study offers key value propositions for architecture, database administration, data governance, data models and testing, infrastructure, integration and UIs. Like operations research and data visualization professionals and data scientists, challenges in this area can be similar, and these professionals are often not familiar with the analytics model development—the nondeterministic aspect—lifecycle.
Organization change management
Analytics-savvy organization change management practitioners are well suited for addressing the specific challenges in this area because of their skill set for working collaboratively with senior change agents in the enterprise. In his Analytics Without Action blog, Seth Godin offers two key points for this field:
- “Don’t measure anything unless the data helps you make a better decision or change your actions.”
- “If you’re not prepared to change your diet or your workouts, don’t get on the scale.”
This messaging aligns with the global business executives study findings by the MIT Sloan Management Review: “The analytics adoption barriers that organizations face most are managerial and cultural, rather than related to data and technology.”
Guiding an organization through the analytics maturity curve is a key value proposition for this area. Booz Allen defines the set of key value propositions to include acceptance, adoption, awareness, ownership and understanding. On the flip side, challenges in this field are often unfamiliarity with operational research or statistical methods of problem solving and the nondeterministic aspect of the analytics model development lifecycle.
IT project management
IT project managers tend to be the unsung heroes of any large transformation program, often taking most of the blame and none of the glory. However, understanding the specific nature of managing analytics projects from traditional IT projects is very important. Professor Stijn Viaene from Katholieke Universiteit Leuven in Belgium categorizes the most important qualities of analytics project managers into five areas:
- Having a delivery orientation and a bias toward execution
- Seeing the value of use and the value of learning
- Working to gain commitment
- Relying on intelligent experimentation
- Promoting smart use of information technology
Key value propositions for this group are statements of work and contracts, project schedules, financial and resource planning, progress reports, risks and action logs. In terms of challenges for this group, they are similar to those faced by organizational change management professionals.
Resources and structure
How should we categorize resources and structure high-performing analytics delivery teams? The answer requires a lengthy discussion and perhaps a dedicated blog. Nonetheless, here are some high-level suggestions:
- Based on education and relevant experiences, each professional should decide on a major and minor course of study from among the nine disciplines defined here. In addition, they need to have significant depth in the major discipline, and reasonable breadth in the minor discipline.
- Having multiple majors and minors is certainly possible. For example, advanced analytics delivery leaders may categorize themselves as pursuing a major in business analysis for supply chain and human resources (HR) and operations research and a minor in finance, project management and information management.
Although we find ourselves in the era of Analytics 3.0, this stage is still early for mass adoption. We do not need to have a winner-take-all attitude. Instead, we need to stay true to our major and minor areas of focus, while continuing to learn in this highly dynamic analytics profession.
Big data analytics professionals will benefit from attending Hadoop Summit 2015, June 9-11, in San Jose, California. We encourage you to come meet with IBM at this exciting, content-rich industry forum. A tremendous learning and networking experience for analytics professionals going deeper into big data, the Hadoop Summit features a great collection of hardcore technical use cases, as well as tips and tricks from many of the most influential thought leaders in the big data analytics industry. Register for Hadoop Summit 2015 today.