A couple of weeks ago I was at two different business schools as a guest speaker. Both ISB, Hyderabad, and MISB Bocconi have specialized programs in business analytics and data science for working professionals. I gave talks about ‘career in data science & industry’s expectations from data scientists’. I got the opportunity to speak to more than 100 students / aspiring data scientists who have gone through the rigorous academic programs in data science. They were all just a few weeks away from their graduation.
First of all, I was really happy to see the enthusiasm among these experienced professionals to make their career transition to my field. I know from my first-hand experience that data science is one of the best professions to be in. For me, it is a sheer joy to find insights hiding in data.
All the students attending these programs were working professionals with an average work experience of close to 9 years. Now the common theme of questions from students was on the following lines
| 1. Do these experienced professionals of other fields need to start afresh in their careers in data science like interns or freshers?
2. What is a good strategy for these professionals to start their careers in data science? 3. How will the industry perceive these professionals? Do they need to take a salary cut while making this transition? 4. How to prepare for data science job interviews? |
As for the last question about data science job interviews, you may like to read the following articles
- Data Science Job Interview Questions and Preparation Strategy
- 7 Worst Mistakes in Data Science Interviews
In this article, I will tackle the first three questions about career transition to data science. Also, I plan to suggest a few pointers for a career transition to data science.
However, the objective of this article is to start a conversation among working professions who are planning a career transition to data science. Please post your questions, suggestions, experience in the comments section at the bottom of this article. I will also participate in this conversation to share my own experience based on your comments.
Career Transition to Data Science and Business Analytics
Sky and Water I – by MC Escher
Let us accept that a career transition, like any other transformation, is not an easy process. This adjacent artwork by my favorite artist, MC Escher, captures the confusion of transformation (in the middle of the picture). Hence, it is no wonder that most experienced professionals making a career transition to data science from other fields are finding it full of confusion and difficulties. Luckily, data science is an extremely vast field with ample scope for most professionals to fit in. This allows professionals with distinct skills to find their niche within data science. I recommend you read the following articles to explore the wide varieties of skill-sets required in data science and ways to polish those skills: Career in Data Science- Play to Your Strength, & 6 Worst Mistakes for Data Scientists.
Let me share my 3 pointers that I believe will help you with your career transition to data science. I will also try to answer the 3 questions raised earlier in the article.
1. Managers & Leaders with Data Science Expertise
Data science is a relatively new field and is evolving extremely fast. Initially, data science teams used to have mostly specialized professionals with skills in number crunching and data management. Over the years problem statements for data science tasks have become much more complicated and interesting. This required multidisciplinary teams to come together and work on data science and business analytics projects. Business teams often work very closely with most of the high visibility and high impact data science projects. Domain experts are always an integral part of every data science project.
Additionally, for most competitive organizations leadership and board meetings are driven by data and facts. This trend is certainly pointing towards a future where managers and leaders need to have proficiency in data science to grow in their career. Now, there will be a direct implication of this for professional investing their effort in learning data science. These professionals considering a transition to data science can build data analysis as a powerful skill on top of their existing skill-set.
Hence, let me come back to the first question I promised to answer at the beginning of this article:
| 1. Do these experienced professionals of other fields need to start afresh in their careers in data science like interns or freshers? |
My answer : no. I don’t think any experienced professional considering a career change to data science should think about starting as an intern. The idea is to add a new powerful skill on top of your existing skill set.
2. Look Around in Your Current Role
Now, let me try to answer the second question i.e.
| 2. What is a good strategy for these professionals to start their careers in data science? |
I believe, a good strategy is to find / create a data science project in your current role and organization. This will certainly add a lot of value to your CV. Most professionals I have interacted with are considering a career / job shift when they are planning their transition to data science. This makes the career transition inorganic and I believe relatively painful. An organic career transition to data science is to add scientific and data-driven thinking in your current role. As I mentioned earlier soon every manager and leader are required to have skills and proficiency in data science, you can be the first generation of leaders in the same direction. In certain cases, you may still have to change your job / career to fulfill your data science dream but I suggest do this only after you have exhausted all the other options.
3. Find Your Purpose
Finally, let’s come to the third and last question:
| 3. How will the industry perceive these professionals? Do they need to take a salary cut while making this transition? |
As I mentioned data science is a rapidly evolving field. The problem statements in data science are becoming highly multidisciplinary. This leaves a lot of room for professionals with different skill sets to collaborate and learn from each other. When we think of core data science skills it still involves quantitative, number crunching, and large data handling skills. However, the projects, that I and most data scientists work on, require creative thinkers with some domain expertise who can pose interesting and innovative business questions. I can tell you from experience, a professional with skills in defining innovative questions and solution approaches always sits on the top of data science hierarchy. These are usually the highest paid professionals. This is also the space where there is a dearth of talent. Now, for experienced professionals, this is the most lucrative space to fill.
Sign-off Note
These were my initial views on career transition in data science for experienced professionals. As mentioned the purpose of this article is to start a conversation between all of us to come up with appropriate strategies. So, please share your questions, suggestions, experience in the comments section at the bottom of this article.
You certainly right about getting experience on data science project. But let us be pragmatic on the industry situation. Projects are not created by someone’s wish and mostly driven by the business need and it is rare when you get actual project with predefined business benefits. And, we end up working on some PoC. Now, it depends that how to take forward the POC and convert into some business benefits. I think this the main problem with Data Science as a career.
I also want to add one more point for experienced professionals is to get adjusted in one of the phase of the data science life cycle and then eventually move towards other components. Like if I am a Data integration lead then I would certainly look at the Data wrangling as my landing area and then eventually start looking into EDA, descriptive statistics and inferential stat. Mathematics will be the next stepping stone when I understand the stat.
What do you think ?
Manish, I think you have made some very relevant and important points in your comment. I completely agree with your latter point about making a gradual transition towards advanced mathematics especially if that is not one’s core strength like for a DI lead. This will make the transition smooth and less strenuous.
Your first point on creation of data science projects let me put it in a different way. The foremost expectation from an experienced professional is that they understand their field reasonably well including the business needs and expected business benefits. Hence, they are better suited to enunciate the business needs and expected benefits than any one else – for instance a banker understands banking and daily business challenges around it. This will also help them sell their data science project to their management. The idea is not to force fit data science on top of an irrelevant business problem but to start with important business questions. Data science or business analytics will serve as a tool to address these problems.
I missed addressing your point about PoCs. To explain to a broader audience, a PoC (Proof of Concept) is the path which most IT and consulting companies take to start / create a project with their clients. The idea is to showcase a quick value of their product / capabilities in a short duration to win a larger engagement with the client to generate a long term value.
In my experience a successful PoC is always driven by business needs and purpose. The main customer for successful data science PoCs is usually business teams (risk, marketing, operations, products etc.). I see a strong trend where well defined problem statements for data science (like plain vanilla scorecards or propensity models) will start to dry very soon. The new trend for data science is about working closely with clients in a consultative model. This requires data science PoCs to focus on problem discovery or business challenge exploration in association with clients. As I have mentioned in the article, this is where data science needs creative and experienced professional with skills in defining innovative questions and solution approaches.
I found your article very helpful Roopam. I am one the students from the Bocconi program who recently graduated. My question to you is this- I have worked in Data Management on ETL side which is one of the key steps in any data analytics project. Would it be right to say that looking to pursue a career in analytics is not a career transition but as in your words” adding more powerful skills to existing skill set”. Hence can I look out for roles in which I can manage data analytics projects or clients?
Hi Meghana, since you have experience in data management and ETL you already understand an important part of data science. However, I think before you move to customer management you may want to spend some time working as a business analyst, and solve a few data science problems. It will help you immensely to be a part of at least a couple of business analytics project life cycles before you manage an entire project on your own. Customer management in data science will test your consultative and advanced problem solving skills.
Can you give me coding for hybrid model for time series forecasting in R. eg ARIMA+ANN
Here is my two cents:
“Data Scientist is the sexiest job of the decade”, “Data Science is currently one of the highest paying job categories”. If these are the main motivations OR the only motivations for the transition to data science, one might be in for disappointment, and after a couple of years might want to change to a new hotter field. I don’t see anything wrong with these motivations, but one should be aware of the consequences.
On the other hand, if you already have inborn affinity to numbers, and have an unabated curiosity to learn, you will find it easier to work in data science related projects and have a satisfying career in this field. You can achieve this without the formal job title of ‘Data scientist’.
Good point Ram. I agree a successful career transition / change to any field is usually not solely based on the reason you have described above. As we all know life is too short to follow one bandwagon after another without a deeper purpose. However, a lot of times most of us don’t know our deeper purpose from the day one, and it is explored slowly. Hence I think one must be open to exploration. The idea is to learn fast (but not in haste) if the field excites you enough to spend rest of your life working on it.
Great point. I would extend he same logic to the basis for a data science project or pitch to a customer too. What I mean by that is just the revenue cannot be a trigger for a data science pitch which may forcefully bring in a need for data science to solve a non problem but its the other way round. One should find a real business problem to solve which will take the engagement with the customer a long way.
Hi Roopam, Thanks for the writing. I am a pre-sales cum solutions engineer by profession and holds 9+ years of experience in IT space. I am looking for a career transition to DS and been looking at different blogs on how to.
I believe that I am strong at Understanding (business side), Analysis (functional), Reasoning (presentation), Questioning (out of box / untapped).
I was good at Math in my Academia but lost touch, but confident on come over.
I am ready to invest my time on learning Things of DS so that I can get to a role where I can orchestrate.
Personally, I love to have a badge of Scientist after my name.
So, please give me a piece of advice on how to start and choose the path especially for a professional with Pre-sales background.
I think for you to learn and implement data science solutions you must consider moving from pre-sales to data science delivery for at least a couple of years. This way you can live a complete life-cycle of a data science project which is not possible in pre-sales.
Within your role in pre-sales I think the best option for you is to get involved in designing a data science solution along with experts followed by an association in PoC / implementation. Learning a new and complicated field from scratch in pre-sales is next to impossible because of other responsibilities in the job. I suggest you move to delivery.
Dear Rupam,
Thanks for sharing the article. How can one find/create data science project within Current role?
Identify a business problem that could be solved using data – this is where your business experience and domain expertise will play a key role. In essence, data provide novel business insights for most, if not all, business problems. The challenge is with availability of data and that’s where you need to scope the problem statement appropriately. Of course, you will have to sell this idea to your boss/company to start things.
After that you are on your own with your creativity and analytical skills. You will have to read and learn a lot during this period. If you enjoy this process then you have kick-started your business analytics and data science career within the secured environment of your current job. This is also a great way to dipstick whether data science is a right field for you.
hi,
I am working as a data analyst for the last 3 years, skills which i gain are SQL and EXCEL, along with this i have learnt R and python.
But companies are asking for the exp in R and Python, how to handle that situation.
shall I go for any part time MS program or kaggle for exp in data science.
Hi Roopam
A brief about me- MBA, 1 year experience in research, 5 months experience in marketing(currently working in this profile), have enrolle in data science course. I’ve tried to come up with a data science work in my current organisation but failed. Recently have across of an offer of unpaid internship in data analyst profile. I feel like I’m standing on crossroads should I go with it or not? Will appreciate your quick response.
It is hard for me to give a clear answer with this limited information. But honestly the answers are for you to explore since it is your career. I could offer some guiding principles those might help you make your decision.
1) I think, when you are thinking in terms of either a fixed job or an unpaid internship, you are getting into a narrow frame of decision making. Think broad i.e. what will it take for you to do both at the same time. You best option is to do both at the same time.
2) Why is the internship unpaid? I have never seen a professional internship with value that doesn’t pay a stipend to the interns. You may want to ask this question to the people you are going to work with.
3) What is the goal of this internship? What will you get out of it? How will it help you? Have clear answers to these questions before you quit your job.
The best option is to do both, if possible. You will learn a valuable lesson in time management. All the best!
Hi Roopam,
Thank you for sharing this Blogs.Need your Help!!!!
I’m currently working in Infosys with 6month experience.My client is British Telecom where I deals with an Application called SMARTS which monitor Network Devices. I want to switch from Infosys to any other data Science Company. I am learning Data Science Concept after work. i wanted to ask you the following things
1. how should I manage to secure a job in Data Science?
2 . Is it okay to lie in a Interview that I worked in Data Science in Infosys?
3. What should I do to master in Data Science(currently do some courses in data Science and improving Programming Skills)?
4. Tricks and tips for cracking Data Science Interviews?
5. What things should I do that will help me to be good Data Scientist as I’m very new to this ?
6. Which company is best for data Science where I can apply for the jobs?
Thank you
Mohammad Rafique