The job responsibilities of a citizen data scientist include dealing with new data, using automated tools to process big data, and creating additional models to gain additional insights. Their primary job is not to make predictions directly from big data, nor develop prescriptive analytics, but to build models and use tools that accomplish those goals.
Citizen data scientists bridge the gap between “true” data scientists (trained and with a degree) and business owners performing their own self-service analytics. This analogy may provide some insight: A data scientist might be able to run ten miles in an hour, but a citizen data scientist can putter around, warm up the car, and drive ten miles in less than an hour, for less money. Granted, the citizen data scientist won’t see as much scenery en route, but they’ll still get the job done.
The position of citizen data scientist is especially unusual, in that, at least for the present, it can only be accessed through in-house promotions. Although the title has existed for a few years, there are no employment listings for employers seeking a “citizen data scientist.” Generally speaking, the position adds responsibilities to someone’s current job description. Getting the promotion will typically involve taking and passing certain Data Science classes that are pertinent to the organization’s needs, and may include a certification.
The creation of the “citizen data scientist” position is a solution to the shortage of data scientists. Much of the work typically done by data scientists deals with mundane operational tasks, such as validating Data Quality, merging data sets, and identifying data sources. These tasks are tedious and time consuming, and having an “expensive” data scientist performing them is not terribly cost effective. It is better to use someone much less expensive in accomplishing these tasks with the help of automation.
According to Ryohei Fujimaki, the founder and CEO of DotData, a software company specializing in automating Data Science:
“Model development, as well as model operationalization, can be significantly simplified by automation. New data science automation platforms will enable enterprises to deploy, operate, and maintain data science processes in production, with minimal efforts, helping companies maximize their AI and ML investments, and their current data team.”
Negotiating the Citizen Data Scientist Position
Management has decided to hire a data scientist for a short-term project and to reorganize the internet sales department. It has also been decided that a permanent “team member” will be assigned to assist the data scientist on a part time basis, as a way to cut costs and retain experience. At the end of the project, the team member will take on the daily maintenance of the newly installed analytics program and predictive algorithms for internet sales. Additionally, the team member will have to take four classes to gain a basic knowledge of the new responsibilities. (A smart, assertive team member might approach management with the idea of being promoted to a citizen data scientist.)
In the situation described above, a large number of changes are taking place within the organization, and unless management is communicating thoroughly to the staff as a whole, there will be confusion and broken expectations. Ideally, the team member will have some responsibilities shifted to other staff. The “chosen one” should also get some time during the work week for studying or attending an online class. The team member should also be involved in choosing the classes, as some online classes are better fits for certain individuals. And then there is the issue of getting a raise. Finally, some sort of arrangement will have to be reached so the newly trained team member doesn’t leave for a new job six months after being trained and promoted.
It should be noted, there might be advantages to having more than one citizen data scientist on staff.
For Management: Creating a Citizen Data Scientist
Selecting the right person is also important. Keep an eye out for people who enjoy reading. In terms of studying, they will have a significant advantage over people who find reading to be boring. Age can be an interesting issue, in that some older people don’t like to learn new tricks, while others may be taking classes on their own, to both continue the learning process and maintain a more flexible mindset.
Providing the right training and tools is especially important when creating a citizen data scientist role. Having decided to make changes in the organization, and to spend money paying for classes, etc., it would be foolish to short circuit the process with poor training and tools that don’t quite work right. Today’s business intelligence and analytics tools combined with an efficient citizen data scientist can help businesses to significantly accelerate their Data Strategy, and improve their profits.
For Current Employees: Becoming a Citizen Data Scientist
For a person with a genuine interest in Data Science, but who can’t return to school full time to gain an advanced degree, the position of citizen data scientist could turn out to be ideal, and a certification program can provide useful training. There are a variety of paths that can be taken, ranging from a self-study process to in-house training to night classes at the local community college. It will depend on the team member’s current skills, the needs of the organization, and selecting the learning approaches that work best for this individual.
It is generally recognized that there are different learning styles and techniques, and that different people learn more quickly and easily with a specific mix of styles. Everyone has a different mix of preferred learning styles. The most basic learning styles are:
- Visual Learning: This kind of student uses images, spatial understanding, and pictures to learn. Students can visualize information easily, and often have a very good sense of direction. The use of whiteboards (or PowerPoint presentations) can be quite effective for this kind of learner.
- Verbal Learning: This type of person learns well through listening, and through discussions. Audio tapes work well. Verbal learners often have large vocabularies and excel in activities that involve speaking, debating, and journalism.
- Physical Learning: These students use their sense of touch to learn. Physical activities are where they excel. These learners enjoy tinkering and learn best when they can do things hands-on, rather than viewing or listening.
A second decision is whether or not to study alone. Some people prefer to study alone, while others prefer to study with a group.
Citizen Data Science Studies
Many places offer online courses designed to provide the basic skills a citizen data scientist needs. There is a very strong probability a student’s employer would add a few classes specific to the needs of the organization, but taking a citizen data science course provides a good foundation. The training should include the following as a starting point:
- Using SQL to prepare data
- Understanding the basic concepts of classification models
- Constructing customer dashboards
- Using SQL to create a segmentation model
- Building a targeting model with machine learning
- Building a recommendation system with machine learning
The Future of Citizen Data Scientists
More and more, organizations are prioritizing the shift to advanced predictive and prescriptive analytics. Currently, traditional data scientists are often expensive and hard to come by. Citizen data scientists can be a very effective way of dealing with this shortage. Technology is the key reason supporting the rise of citizen data scientists. Technology has made it easier for non-specialists to accomplish the same goals. In the last few years, Analytics and BI tools have become significantly easier to work with and include augmented analytics.
Image used under license from Shutterstock.com