Why should you care what was your team members use?
The first, of course is that your team. If you think you can help one of your teammates do better, it certainly seems like good sportsmanship (or whatever the equivalent is in the business world) to at least make the suggestion. Even more directly, the fact that your team means that you will probably have to deal with some of the work that they put out. It’s far easier to build upon something in your tool chains are sufficiently aligned. Of course, even getting more similar tools has huge value over using very different tools and techniques.
Picture this: you have been with updating some work that was done over a year ago. The person who did that work is on vacation, but they sent you their final presentation, and a link to the shared directory where they kept all of their analysis. You go To the directory and see 12 different Excel files, with names differing only in the cryptic appendage of numbers and “final version” (used on four of the files). Looking through the presentation, you see one surprising finding, which seems too good to be true. You decide to start digging around in the Excel file to try to understand where that number came from. Sorting by last modified date, you start picking through a file trying to find where that number came from. Pretty soon, you have narrowed your goals to understanding even the basic data flow. When you think you have that, you load in updated raw data into it refresh. Everything breaks.
Why should you suggest Python?
Strictly speaking, I can’t guarantee that Python will be the best for your organization. But it’s critical to make at least some concrete suggestion, and get as many people on the same page as possible. If there is a compelling argument for using R, SAS, STATA, or any of the newer languages, those are probably the way to go. But absent any situation specific concerns, Python is what I generally recommend, particularly for teams which haven’t done much of their analysis with code before.
One huge benefit of Python is the community. Because it’s a general-purpose programming language, in addition to being extremely useful for data analysis, there are thousands of people online who have answered every conceivable question. The popular programming website stack overflow, which lets people ask questions for the crowd and vote on the best answer, has more than 500,000 python questions. That’s more than five times the number of R questions, and six times the number of Excel questions. So I can almost guarantee you that any issues you run into in your first few months of basic exploration has been extensively covered and are just a Google search away.
Once you want to start using the tool for more specific work, you can leverage the huge variety of books that are available on Python. Many of them are free, and those that are not very easily worth the $20 or $30 that they may cost. Remember the value of your time, remember how much of it is spent manually searching through spreadsheets! http://importpython.com/books/ keeps a good list of the options.
Great, what should I do now?
Fortunately, there are a bunch of great ways to get started with Python quickly. And I mean really quickly, like you-don’t-even-have-to-install-it quickly. In fact, you’ve wasted too much time reading the paragraph already. Go here: Wakari
Hosting the analysis of your company’s data may not be feasible, but these instant on solutions are perfect for giving people a taste of what-based analysis can do. Once they begin to visualize how these tools can improve the analyses that they’re currently working on, it will be much easier to convince them to walk through the local setup process!