Contrary to what you may have heard, an analyst and a data scientist are two separate entities. From a career standpoint, the roles are more complementary instead of one position being at a higher level than the other. However, if you want to become a data scientist, and that is a personal goal for you, then you’ve come to the right place. Keep in mind that it’s also possible to make the switch from data scientist to analyst. Either way, I’ll go over the roles and responsibilities of your typical analyst and data scientist, skills you may need, and how to transition specifically from an analyst to data scientist.
The Analyst Role
First, we’ll dive into the main purpose of an analyst. The main focus of an analyst is to understand the business, use data to help measure how the business is performing, and help drive decisions regarding that business. Before burying themselves in data though, business context is extremely important to understand. It is critical to understand the goals the business is trying to accomplish.
Regarding tools, an analyst shouldn’t be defined in terms of an “Adobe Analytics” expert or “MySQL” expert. The tools they use are definitely important, but an analyst is someone who can primarily look at any set of data from any tool and derive meaning from that set of data. They go beyond the “how do I use this tool” and instead focus on answering “what is happening” and “why.”
In many cases, an analyst may not know exactly why something is happening. Analysts are constantly generating a lot of questions but may not necessarily have the answer to all of them. What they can do though is give some possible hypotheses, provide data to support them, and then validate and test those hypotheses.
What Makes a Great Analyst
There are certain characteristics and skills needed to be a great analyst.
- Be a critical thinker.
- Be investigative and creative. You can’t just stop at the first step, but must take a simple question and provide deep insight into it.
- Understand how data is collected.
- Analyze any set of data and derive meaning from that data.
- Identify problematic issues surrounding data.
- Multitask and analyze multiple data sets while potentially cross-comparing those data sets.
- Be skilled at presenting any findings.
The Data Scientist Role
There is a significant overlap between a data scientist and an analyst. Data scientists also require business context and any information regarding business goals. The process of a data scientist begins with asking what they are trying to answer based on what they know about the business.
However, where they start to diverge is the technical skills required to be a data scientist. Data scientists get into the details of the “why” by using calculations to make predictions. They are more future-oriented and focused on discovering what will happen and what the business can do about it based on their findings from various scientific methods.
Scientific methods a data scientist may use include linear regression, decision trees, and neural networks. From their methods, they will disseminate that information so any role in or outside the business can understand the data findings. If a data scientist is doing predictive models, they may use an automated system to display it instead of presenting their findings themselves.
Deep technical/computer science skills is also something that a data scientist needs. Since building predictive and prescriptive models typically require accessing multiple data sets, applying complex statistical methods to such data sets, and delivering the results in an automated fashion, software development skills are critical. However, it is important to note that software engineering skills do not equate to data scientist skills. You can be great at using and developing all the data science tools in the world, but if you don’t properly understand the business problem and the underlying methods to arrive at an answer, you could cause more harm than good!
What Makes a Great Data Scientist
A good portion of what an analyst does, a data scientist needs to have similar skills, but there are some key differences.
- Be a critical thinker.
- Be investigative and creative.
- Have a comprehensive, deep-level understanding of statistical methods and experimenting.
- Be skilled in predictive analytics.
- Be future-oriented.
- Have deep technical knowledge and acumen.
Making the Transition from Analyst to Data Scientist
If you’re an analyst, you shouldn’t feel pressured to become a data scientist. It’s not a linear progression where you start as an analyst and work your way up to a data scientist. From a team perspective, a good data scientist can benefit from having a good analyst on the team. Conversely, a good analyst can also benefit from having a good data scientist on the team. There’s overlap between the two roles, but they certainly can co-exist . However, if you are looking for a change or becoming a data scientist is a goal of yours, these practical steps and resources can help get you there.
- Hit the books on math and statistics. Learn the fundamentals of math and statistics. It is critical to understand the underlying calculations that the tools and libraries are performing behind the scenes before applying them to problems.
- Learn the tools of the trade. Learn some of the different methods and techniques, like the previously mentioned linear regression, decision trees, and neural networks. If you understand the math and statistics from step 1, you’ll better understand the various pros and cons with each of those methods and techniques.
- Learn the technology. Once you have math and statistics under your belt, along with the various methods and tools needed, learn to code really well. A couple popular programming languages for data science are Python or R.
Here are some resources to help you on your journey to become a data scientist:
- Coursera Data Science Specialization (https://www.coursera.org/specializations/jhu-data-science)
- General Assembly Data Science Course (https://generalassemb.ly/education/data-science)
- Research university courses online or see what’s available in your community
By implementing the steps above, you’ll be off to writing models that deliver results you can share with your business, stating what you think will happen. Over time you can measure your predictions and grow your technical skills so you can work with your development team to integrate your predictions and optimize digital properties.
You’re a Data Scientist, Now What?
Making the transition from analyst to data scientist is not an easy task, but when you do get there, here are a few tips to keep in mind.
- Being a data scientist is not a destination. Technically, there is no such thing as a data scientist who has “arrived.” You will continue to learn and experiment, while embracing your role as a new lifestyle.
- The goal isn’t to be always right. Your methods will not always be the most correct methods. Make your predictions better and improve by continuously measuring over time.
- Don’t try to be a unicorn. It’s OK to ask for help. Work with other people who know what they are doing, especially if you don’t understand something. For example, if you have a data set that’s hard to reach or requires a lot of data wrangling, work with data engineers if you have them. If your presentation skills are lacking, try working with an analyst.
The roles of an analyst and data scientist are two separate entities. One is not better than the other, but they do have different focuses. If you are set on making the transition from an analyst to a data scientist, you may initially feel overwhelmed with the amount you need to learn. My best piece of advice is to learn according to your style. Leverage your strengths and learn the way that works best for you. And always remember, don’t be a unicorn. Reach out for help when you need it.