Suppose one wants to work as a data scientist. How important it is to learn Excel or Libre office programming (macros), or is it enough to learn R or Python well?

My question is, how should someone determine what skills are relevant to focus on learning, when preparing for a new career?

  • @Kaz There are literally hundreds of viable technology choices for data analysis. We can not really make any reliable recommendations in this regard. – Philipp Feb 14 '20 at 16:16
  • Not specific recommendations, no. But seeing as the OP doesn’t seem to have a clear idea themselves what they want to do yet, I’d recommend Python as the best (general) one to learn as it opens up the most options. – Kaz Feb 14 '20 at 17:27
  • Should a carpenter know how to use a utility knife? Excel / Calc is your utility knife. Basic proficiency will help you finish each day's work faster. – O. Jones Feb 16 '20 at 21:09

Anyone doing data analytics should know Excel, because no matter what else you learn, the information is going to need to be presented.

Excel can connect to virtually every data source out there, and knowing the formulas, VBA, and the C# to create xlsa files will put you in demand.

20 years ago, there was a concerted effort to expunge Excel from the workplace, it is so entrenched, and so ubiquitous and at the same time, so under utilized that knowledge of it gives you a leg up in the industry.


I have been working with Excel for 20 year now. It's everywhere, not going away, and an immense asset to anyone who knows how to use it well.

  • The problem with questions like this is that there are often exceptions and a whole spectrum of realities at different companies. I run a team of BI developers and data analysts and we haven't used Excel to present data or share findings in years. Much less to store or manipulate data. It's installed on my laptop but I can't remember the last time I opened it. When we need to present, it's in PowerBI, Tableau, SSRS, or other enterprise tools. I would happily hire a developer who knew SQL, Python, R, and PowerBI, but not Excel. The expunging of Excel did actually happen in some employers. – dwizum Feb 12 '20 at 22:10
  • @dwizum yes, but we don't answer questions based on exceptions, there are too many exceptions – Old_Lamplighter Feb 12 '20 at 22:16
  • 4
    I have to occasionally present results in Excel but these are produced in R and written into Excel with some suitable R libraries. This also makes them a lot more reproducable and adjustable than having formulas directly in Excel. In my company, in spite of results being requested in Excel doing any computation in Excel is considered bad style and should be avoided. – quarague Feb 13 '20 at 7:52
  • 1
    @dwizum Oh, I get what you're saying. Maybe I should clarify that of all the tools, it has the widest use. If I was looking for a web developer, I'd WANT someone to be able to use a text editor if need be (you'd be surprised how often you need to revert back to basic tools) – Old_Lamplighter Feb 13 '20 at 15:53
  • 1
    @dwizum I voted to reopen, I'll post to chat asking for a fifth vote – Old_Lamplighter Feb 14 '20 at 3:20

From my experience I would say learning as much as you can in all systems related to your work.

If you work heavily in Excel then formulas and VBA will help immensely. If you are working with raw data from some database(s) then using a well documented language like python and useful libraries like numpy and matplotlib will be massive tools to work towards improving your skill set in this field.

As far as your question:

Is it enough?

I would say no if you plan on being proficient and a person people will go to for getting things done.

VBA/Formulas are a good start but for good data analysts you will want to familiarize yourself with SQL, Python, useful Python libraries, Other popular data science languages like R and visualization tools. IE: Tableau, looker and so on.

  • 1
    Yep, learn as much as you can, each have strengths – Kilisi Feb 13 '20 at 3:24

The core of your question is how to determine what skills are important for a new career. Luckily, the answer to that remains the same, regardless of the career:

  • Do your research. Look at job postings in your desired career path. Don't look just for jobs you actually want to apply to - look for jobs more senior (and less senior) as well. Write down a list of key words and requirements.
  • Similarly, look through LinkedIn or other professional networking tools for people who already have those jobs. Find the profiles of people doing the work you want to do. Write down the skills those people have, and check their work histories and educational histories for any trends. See what their stories are.
  • Look for online forums or discussion groups that focus specifically on the career path you're interested in. See what tools get mentioned most often.
  • Perhaps most importantly, consider the nuanced differences in the positions and employees you're researching. People with the same title, but working in different environments, may have very different skill sets. You may think this will make your research less valuable, but in reality, it can be the most valuable part of your research, because it will ultimately be what helps you answer your question. More on this later.

It sounds like you've already done some of that baseline research, because you've arrived at a short list of technologies you're thinking of learning more about. But, before we actually address your specific problem of which tools to learn, you need to do some more thinking about what your ultimate goals are.

Ultimately, many problems can be solved with many tools, and there may be subtle differences between the different solutions. Some employers will care about these differences, others may only care about getting the work done. If you're starting from scratch, and will be looking for entry level jobs anyways, it may be just as valuable to show that you can pick a technology and learn it, versus trying to promote yourself as already being an expert in a specific technology. With this in mind, if you're in an interview, be thoughtful about talking through your problem solving approach and your approach to learning new things versus trying to focus on describing how you use or prefer specific tools.

In fact, focusing too hard on specific tools can cut both ways. If you happen to focus on a tool that a specific employer likes, it may be helpful to describe your expertise in it. But, focusing too much on a tool that an employer doesn't prefer can disqualify you. Since you've asked specifically about Excel, it's important to consider a few things: Programming in Excel can be incredibly powerful. You can solve many problems with it. In a general professional office environment, people who are highly skilled in Excel will be seen as Problem Solving Gods.

However, Excel doesn't scale well to true data science problem solving, and it usually isn't preferred as a primary tool among focused data science teams working on large and complex problems.

This brings us back to my bolded comment above. Once you have your baseline for skills, you can refine and focus in on the skills preferred for the type of environment you want to work in:

  • If you want to be the lone data analyst at a community bank for instance, being an expert in Excel is probably a great choice. You'll be surrounded by Excel users every day, and you'll be working with problems that it's great at solving. Your boss in such an environment may not know anything about Python, but they will certainly recognize Excel.
  • If you want to join a larger organization with a focused team, or one that is totally focused on data science, trying to show off your Excel skills during an interview may get you odd looks at best. But - talking about how you went to the useR 2020 conference and you really enjoyed the tutorial on integrating R with C++ will get you hired.

or is it enough to learn R or Python well?

Actually, the question should be the other way round: Is VBA (macros) enough or is Python/ R needed?

Data scientist and data analyst can mean everything depending on the company.

But generalizing, macros can be helpful if you are starting as a data analyst and are to take over such tasks as automating reporting and some basic analyses.

But Python or R are necessary for more complex stuff.


How important it is to learn Excel or Libre office programming (macros), or is it enough to learn R or Python well?

If I were writing the job description, I'd put it under the category of "nice to have" but certainly not a deal breaker if you don't. As others have pointed out, Excel is an incredibly versatile tool that can be used for everything from simple data entry to building complex forecasting models. One of the advantages it has over programming languages like R or Python is that it's quick and easy to set up an interface where users can enter data and make adjustments to calculations as needed. This is probably not an optimal solution in the long term but when there's a tight deadline to get a process up and running and you don't have the resources to build a custom UI, a spreadsheet with VBA macros starts to look like a very attractive option.

The reason I say it shouldn't be a deal breaker is that data science work is mostly theoretical. If you understand statistics, linear algebra, and algorithm design, your familiarity with any one particular tool is of secondary importance. In 2020, most data science teams use either R or Python or a combination of the two in their daily workflow so these should probably be your main focus as far as tools are concerned. If you have all of the above, I would trust that you could pick up Excel macros without too much difficulty, should there be a need for it.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .