We can call 21st century as a time of information. Nowadays Google generates more information than all companies in the world did 15 years ago. For example in 2000 world produced about 1.5 exabytes while in 2025 we can expect growth up to 175 zettabytes. Why should humanity store all this source of valuable information at the shelf while we can analyze it? In 2017 Google analyzed data from NASA’s Kepler telescope and found new detected multi-planet star system that houses 8 planets of Earth-like conditions. Vancouver’s police uses
unique program that analyzed data of break-in criminal activity archived from 2012 to 2015 and every two hours makes decision of where next crime to occur. It reminds me about futuristic Robocop movie where the guardian of the law in the shape of semi-man and semi-robot served as a police officer.
These discoveries can not be achieved without two well-know jobs such as data-scientist and data analyst. For the past few years Baruch students became more interested in data related field (especially in two promising jobs: data science or data analyst) full of new job opportunities and often confusion. Let’s take the torch (or may be spotlight?) and go inside of a dark cave hiding unknown secrets, bring them out on the surface and make our unbiased decision of which one is better to pursue.
Advent of the Data Scientist and Data Analyst
Both data science and data analysis belongs to new discipline business analytics first appearing in 1985 with a launch of Microsoft Excel. Before this, all simple business analysis was conducted manually in companies by statisticians. More substantial and influential analysis was rather a work of researchers who had an access to relatively primitive and
at the same time large computers in their labs called mainframes. For example on pic. 2 mainframe in NASA Glen research center is as large as room and does significant research and technology development on jet engines, producing designs that reduce energy consumption, pollution, and noise. Business analytics is a result of two major trends: technology and data growth. First allowed people to work efficiently with electronically documents, easy edit and get quick access. Second is a result of technological boom, cheap data storage produced huge amount of information that can be potentially used for solving business problems.
What do they Do?
In the very beginning of our journey let’s clear up what are the responsibilities of each job. These vacancies appeared gradually in firms since 2013. They are not well defined for many recruiters who don’t understand yet the differences between data jobs and often mixing them up (data analyst, data engineer, data scientist). Even @silentbicycle made a comic statement in his Twitter account that the only difference is that data scientist works in California.
Of course it’s not true. First of all both data science and data analysis are quantitative and require decent amount of mathematics and statistics. Data analyst sift through data and provide reports, charts and diagrams to what the data is hiding. For example, these are typical responsibilities for search “data analyst” in indeed.com, an American worldwide employment-related search engine for job listings.
-
Interpret data, analyze results using statistical techniques and provide ongoing reports
-
Identify, analyze, and interpret trends or patterns in complex data sets
-
Filter and “clean” data by reviewing computer reports, printouts, and performance indicators to locate and correct code problem
At its core, a data scientist’s job is to collect and analyze data, garner actionable insights, and share those insights with their company. While data analyst uses prepared data, data scientist has more wide responsibilities and must be able to do three essential steps: collect, analyze/experiment, share results.
Collection, cleaning, and munging data is first and essential step and it’s required for all unstructured data that constitutes 80%. In the chart it shown with red line, it can be seen that in next few years specialist who can prepare data, will be in high demand. Data scientist are one of them. When data are organized its time to find patterns, build models, and use algorithms. Unlike data analyst who uses relatively primitive technics such as statistics data scientist implement advanced tools such as machine learning. Moreover, data analysts alway know what they are looking for or what analysis they do. The result can be one of many known indicators such as KPI, operation cash flow or working capital. Opposite data scientists usually don’t know what they are looking for. In a complex, disorganized databases there can be a lot of amazing discoveries as results of experiments rather application of know formula. Last but not least difference in responsibility is the necessity to communicate with team members, engineers, and leadership. When project completed data scientist is the only one who knows how to achieve and interpret the result (to the contrary every specialist know how to interpret KPI working capital indicators). Summing up the difference: data analyst does relatively simple statistical analysis while data scientist uses advanced techniques to discover hidden relations in data.
Career Comparison: Business Analyst vs. Data Analyst?
Career path are also different despite the fact that both are quantitative jobs. Data analyst required to have undergraduate degree. Advanced degree is not required. Often companies don’t even ask bachelor if candidate has experience. 88 percent of data scientists hold a master’s degree and 46 percent have a Ph.D. Ideally both jobs require mathematics, statistics, computer science and business knowledges. If data analyst may not be a guru in statics than data scientist must. Besides, no data scientist can go far without deep computer science experience. Famous phrase among data professionals represents that computer science less important though.
One of the main difference between two jobs is that data scientists usually apply machine learning. There is no machine learning in data analysis while in data science all based on machine learning. The word learning in machine learning means that the algorithms depend on some data, used as a training set, to fine-tune some model or algorithm parameters. In other words data scientist build software, train (teach) with specific data and apply to real world information flows. According to Martin Schedlbauer, associate clinical professor and director of Northeastern University’s information, data science, and data analytics programs,
“Data scientists are quite different from data analysts; they’re much more technical and mathematical. They’ll have more of a background in computer science, and most businesses want an advanced degree.”
Browsing internet I came across this amazing Venn diagram from Stephen Kolassa’s (Data Science expert in SAP) post.
It shows that ideal data scientists located right in the middle. It well educated person in several fields and includes statistics, computer science, business and communication. Let’s not forget that they should be experts in statistics and business side. Probably not a best choice to pursue data science after 30, though its a good option for Baruch students. Baruch has solid math/stat department and one of the best business schools in the country. Data analysts in this diagram is “Analyst”, depends on level of computer science knowledge. “Analysts” has communication skills and poor in computer science. Let’s not forget that they have less knowledges in statistic. It seems that data analysts would be a a good start.
Mobility is important
These set of backgrounds lead us to another difference which is mobility. Data Scientist find it much easier to changer their career path if the don’t like it. Mathematics considered one of these subjects that allows alumni to find a job in broad circle of quantitative fields. Additional set of statistics, computer science and business encompassing almost all possible options. While demand on this vacancy is high only growth from year to year
they can choose companies in considered competitive cities such as New York or San Francisco. Data scientists are particularly in demand in Seattle too, as the city faces one of the biggest shortages of data scientists in the U.S. Even Harvard Business Review has named ‘Data scientist’ as the “sexiest job of the 21st century”. Data analysts vacancy came to the market years before and also is in demand. As can be show on a digram below number of data analyst vacancies has steady growth and can hold a “hot job” title too.
It seems like data analyst is a intermediate step to be data scientist and have a hard time switching to another field if they want to. Lack of advanced knowledges in machine learning and statistics pushes them to data science. From another side data scientists are open to work in pretty much any field with data. Their broad experience lets them open a business at the end of the career path.
Comparing annual salary and perks
Before making any conclusion I would like to compare annual salary or job benefits. Since both of them are STEM jobs (shorthand for science, technology, engineering, and math) they both well paid. In the graph provided by Glassdoor – current and former employees anonymously review of companies and their management – data scientists takes leading position among all data related vacancies while data analysts are in the very end. Entry level salary starts from $95,000 and 65,000 respectively for data scientists and data analysts which is still a reasonable investment in college debt regardless of selected career path.
It’s obvious that data scientist overpass data analyst with their almost twice higher salary and includes number of perks. It’s a safe career to pursue. It will not be replaced by artificial intelligence in nearest future just because data scientists produce them and we still don’t know how to create an intellect that would be creative enough to produce another intellects. Although building a business becomes easy. When you know that much inside and out of many industries and when building contacts and gain the ability to solve real-world business problems is goal of your job, it becomes easy to establish your own business. The author, Megan Mary Jane says
“Those who always dream about building their own business before retirement, their experience, contacts, and knowledge as a data scientist can be helpful in their future endeavors.”
Another reason to become a data scientist is unlike many other IT jobs, they do not have to create useless study material for beginners. Many courses available in known online study platforms. One of the leading is coursera.com
They backed by experts with solid experience and knowledge in the field. Unfortunately data analysts can not brag with anything outstanding though they have something that many vacancies don’t. Data analysts is very good point to start a career in the data science. Few years experience gives broad understanding of the related field, practical skills in statistics, mathematics and work with the data. Obviously data scientists are winner in this section: paid more than analysis, the spend less time doing boring work (such as writing tutorials) and as a result can spend more time on experiments, such a nice job!
Summing up data science and data analyst – two very close vacancies – it’s important to highlight a few facts. It’s tough to study for data science since 80% candidates received their master degree and relatively easy to get a job of data analyst with the only requirement of undergraduate diploma. Both are in high demand in the market. Data scientists are paid more and have more perks. These two are usually mixed up by recruiters so be prepared to ask questions about work responsibilities. It seems like data science is a better option for experienced data analyst who would like to do something else and have time to expand skills in several non related field such as computer science, communication, business. Taking into account high level of math and statistic in Baruch it can reasonable for Baruch students to pursue data analysis, in the beginning of their career path and switch to data science when they get necessary experience. Make sure you to get solid background in computer science by that time, it will help you not to give up and get through machine learning course.