Analytics is an evolving field, with new techniques being created on a regular basis. Providing a training plan for analytics professionals assures their continued success, and fortunately there are many options for conferences, meetups, online courses, etc., to meet this need. We will provide information on INFORMS resources, but analytics leaders are also encouraged to explore other options, including those that may be industry-specific. INFORMS provides the Certified Analytics Professional (CAP®) program that sets standards that should be met by analytics professionals. This is supported by the INFORMS Professional Development program, which offers training to enhance skills. Finally, INFORMS provides conferences that allow analytics professionals to network and learn about the latest best practices in analytics.
The kind of talent you hire will be first and foremost a function of the needs of your business and the structure of your organization. Determining this is a fundamental piece of how organizations can get started with advanced analytics. Consider who will “consume” your work – will it be other technical people within a larger team of analytical resources or nontechnical people in the business? If the former (more typical in a larger organization), you may need to focus on technical skills and be able to afford some specialization. If the latter, you will likely need staff who are stronger in “soft skills,” so they know how to communicate technical topics to nontechnical teams.
There are five types of analytics/data science skills to consider in building out the team. You will need all five in your organization, but the type of structure you have will dictate whether you need most of them in one person or if you can afford specialization.
- Data skills – Deriving insights from business requires the ability to pull data out of the organization’s big data repositories and other data sources. Data collection will almost always take place in disparate locations, necessitating skills in data extraction and database querying, joining tables, and data architecture. Being data literate and understanding the content and context of the data is also quite important.
- Programming skills – To have the flexibility needed to perform advanced analytics, programming skills are essential for data manipulation as well as the modeling. Programming languages change over time and vary by discipline, but the necessary characteristic is the comfort in writing reproducible code to perform tasks, the ability to learn new languages as needed, and the skill in documenting code for transparency. Increasingly, new tools catering to the self-service analytics market require minimal or no code and provide self-documenting capabilities.
- Modeling and related quantitative skills – Transforming data into insights requires advanced analytics, but the kind of analytics capabilities needed for your organization will vary based on your organizational needs. A solid foundation in math and statistics is a critical base upon which to build. This provides the analysis with the rigor needed to produce credible results. This foundation is particularly important if the analytics team lives in an organization that has less technical depth. Although modern tools and easy access to programming packages mean it is possible to easily produce results, ensuring that the results are meaningful and practically significant is much harder. Sometimes it is better to have no model than the wrong model, and you need to hire someone who knows the difference. Upon this foundation of statistics and math are individual specialties as described in the sidebar at the beginning of this document.
- Business acumen – For the successful application of data analytics, understanding business processes and how they can be modeled using analytic techniques is a critical skill for success. Knowing the details of the different analytics tools and techniques is important, but the skill of mapping business processes to the appropriate techniques is a requirement for success. This is what separates pure statistics from "business intelligence," which connects statistical conclusions to business cases.
- Soft skills – The ability to clearly communicate technical content in a nontechnical manner is important for analytics success. The analytics enablers are not interested in the latest algorithms, mathematical theorems, or statistical software. They need to understand how applying those tools and techniques to organizational data will provide better decisions for their organization, and ultimately some form of value. Data scientists should be able to write good reports, give clear and concise presentations, and must have good listening skills as well.
An analytics team can be built by hiring full-time, permanent employees or by using contract staff, external consultants, vendors - or some combination of these. Organizations that are small or new to analytics may choose to start by contracting the work out via staffing agencies, outsourcers, consulting firms, or vendors who offer some form of “analytics as a service.” These methods can be a quick way to value in the short-term but do not build internal capacity and result in capacity “leakage” when considerable time is invested in providing domain expertise to the external party for them to be effective.
Building a core analytics team and analytics strategy may initially take longer but develops internal capacity that may be valuable in the long run. Analytical talent is in high demand, which leads to both competition for candidates and a proliferation of those who purport to have analytical skills but may lack the necessary depth to be effective. One way to assess talent is to seek out candidates who are Certified Analytics Professionals, which means they have demonstrated a certain level of competency. If you are starting from scratch to build an analytics competency within your organization, it may be worth hiring a consultant who knows the field to help you get started.
Finding analytical talent is not the same as finding talent for the IT organization, so recruiters from human resources will likely need guidance to appropriately vet candidates. One source of candidates is the INFORMS Career Center, which allows employers to post jobs requiring analytics talent, and job seekers to upload their resumes. In addition, INFORMS organizes a Career Fair at both the INFORMS Annual Meeting held in the fall and the INFORMS Business Analytics Conference held every spring. The Career Fair offers an opportunity for employers to discover a wide range of potential candidates and meet the candidates on-site for one-on-one interviews.