Is Analytics always the same? An industry-based exploration of skills needed for analytics jobs.

By: Dr. Deborah Sliver, Karen Bemis and Sue Weston

Analytics and big data are changing the way we do business, creating a demand for new jobs with different skills. Companies are shifting from generating data to relying on data to make decisions. Are there differences in the skills requested for analytics/data science positions based on the industry? To answer this question, we focused on four sectors: consulting, healthcare, financial services, and insurance. These industries were selected based on our students’ interests and the concentrations of jobs in our region.  We tailored the analysis to focus on positions suitable for MBS students, looking at advertised jobs requiring under five years of experience and a bachelor’s degree. (Information about skills came from Burning Glass. The query included jobs posted in analytics between July 1, 2017, and June 30, 2018, please see the methods section below for a description of the data.)

Who is hiring? There were 67,000 jobs advertised for positions in analytics over the past 12 months. The four industries we targeted accounted for about 40% of these roles. The chart (below) shows some employers.

Jobs by industry (based on % jobs created) and some associated employers.

The CVS Aetna merger which combines healthcare and insurance providers begins to blur industry differences and may be the beginning of other health-care deals such as Cigna's plan to buy Express Scripts

What skills are important? The chart (below) compares the most requested skills. SQL is a must-have skill for analytics across all industries; it is expected to remain at current levels while other skills, like Python, are expected to increase in demand. Project management is required for all sectors and appears on approximately 1/5 of positions. Other recommended skills include SAS and business intelligence. 

Top 10 skills by industry (skills in green are projected to increase in demand)

What skills are different? Healthcare and insurance roles require a similar suite of skills, but the relative importance differs. For example, big data was requested on 22% of insurance jobs but only on 9% of healthcare roles. Healthcare places a premium on business intelligence (36%) and data analysis (29%) whereas both these skills appear in approximately a quarter of roles in healthcare. Python is used less in healthcare than other industries (on 19% of positions compared to 32% in insurance). Both industries include Tableau on about a quarter of jobs. Insurance roles advertise computer skills including PIG (6%), NoSQL and Scrum (5%). Healthcare includes data warehousing for 20% of positions; this competency is expected to decline in demand.

Financial services roles include some business specific requirements such as economics (on 25% of the analytics jobs) and risk management (on 21% of positions). Computer skills identified include Cassandra (3%), Natural Language Processing and VBA (4%). 

Consulting roles specified machine learning and big data more than other industry and showed the lowest demand for data analytics (14% compared to over 24%). Python nudged in front of SQL as the most requested skill. Core competencies include business acumen (17%) and artificial intelligence (15%). Computer skills in demand include Neural Networks (13% of jobs), Natural Language Programming (10%), Cassandra (9%), MongoDB (8%) and PIG (7%). Surprisingly, 6% of jobs requested robotics. Overall, consulting jobs require more programming skills than the other industries.

What are the fastest growing skills across all industries? Employers are expected to increase their need for predictive analysis, machine learning and data visualization (which is why knowledge of Tableau or an equivalent tool is essential).  Burning Glass predicts an increase in demand for more nuanced skills such as NoSQL for data handling and R for data analysis. Other software skills anticipated to increase in desirability include Sqoop, used to import data from relational databases, and Apache Hive / Hadoop for data summarization, query, and analysis.

Where are the jobs? The chart (below) presents the percentage of jobs by region. It shows that our region led the country in jobs for insurance, finance, and healthcare and was second behind Washington for consulting.  Where is California? When we looked across all industries by state California emerged, generating the most jobs (19%) followed by New York (9%).  Metropolitan Statistical Areas (MSA) provides a more granular view, focusing on geographical regions with a high population density at its core and economic ties throughout the area MSAs placed The New York–New Jersey MSA first with 12% of jobs, followed by the San Francisco-Oakland-Hayward, California and Washington-Arlington each with 7% of jobs.

Locations of jobs posted by metropolitan areas (MSA) with our region highlighted in green.

Emerging trends by industry:

Insurance industries will be developing data-driven decisions, which according to a Gartner study, will evolve structured data incorporating voice, video, image and text data. Better use of behavioral economics modeling can reduce insurance fraud while improving fraud prevention. Decision optimization with chatbots can personalize the client experiences for filing claims triggered by using life events creating touchless claims which will enhance productivity. 

Finance industries will be using big data to understand customer loyalty and provide customized experiences. Understanding their customers will support inventory management, improving automatic bill payment or providing customized wealth management solutions. Anti-money laundering and data privacy are other hot topics which depend on data. Machine learning will allow institutions to manage risk. Blockchain technology will improve transaction security dealing with sensitive data. 

Healthcare: Leveraging data will generate personalized solutions to reduce costs and save lives. Wearables provide biometrics which can be used to create two-way connections changing the way companies conduct research. The availability of data will improve predictive models, and big data will provide real-time infection control.   

Consulting: Data analytics will continue to extract information and intelligence essential for understanding the global marketplace. Unstructured and real-time data will provide companies with a competitive advantage and individualize the client experience. 

Data analytics will enable organizations to compete with potential disruption and position them for success.

Related courses offered at Rutgers

Definition of Terms:

Methods: 

This blog analyzes jobs advertised between July 1, 2017, and June 30, 2018, using a tool called Labor Insight from Burning Glass Technologies. By mining the detailed information stored in job postings, we can determine what employers are looking for when they fill analytics positions. The positions were selected based on having a job title containing the words: data scientist, big data, machine learning, data engineer, intelligence or analytics. There may be many more jobs that could be classified as "data science" or "big data jobs" but we focused on those with a title that reflects this new area. This analysis identifies skills predicted to increase in demand based on a methodology developed by Burning Glass. These projections combine econometric time series models with machine learning approaches to predict the growth in the job posting’s demand for skills. We used the Metropolitan Statistical Area (MSA) to define our region. MSA is a geographical region with a high population density at its core and economic ties throughout the area. Three MSAs compose our region: Philadelphia-Camden-Wilmington, New York-Newark-Jersey City, and Trenton.

While this analysis can show trends in the job market, there are limitations. We only included jobs advertised online. The unstructured nature of job ads can make it difficult for the system to identify individual pieces of information effectively in some cases. While Labor Insight breaks up the job description into fields for analysis, inconsistency in the formatting of job descriptions and industry-specific terminology or titles may result in the inclusion of some irrelevant jobs.