I’m not ashamed to admit that I’ve been using spreadsheet software since the 80′s (props to my dad and brother for turning me ontoVisiCalc during the Apple IIe days). I’ve also been in love with the predominant spreadsheet software, Microsoft Excel, more or less since its inception. And as scary as it might sound, I’m not overstating things when I tell you that I eagerly await the arrival of each and every one of Annie Cushing’s wondrous treatises on advanced Excel functionality.
And yet I don’t have Excel on my personal computer.
Why? Because while Excel is certainly a very powerful software application with a wide range built-in functionality as well as a plethora of outstanding bells and whistles, my experiences over the past few years has suggested to me that Excel is simply incapable of handling the immense data sets that are fast becoming the norm in today’s (and tomorrow’s) enterprise business landscape.
I first realized this a few years ago, after building some custom Excel formulas and data tables that helped automate various types of acquisition marketing analysis. The formulas worked like a dream, but as the data sets became larger and larger, the delays, crashes, and file corruptions became a frequent and very painful nuisance. Eventually, it became quite clear that in order to scale our efforts my team and I would have to shift to more powerful programming languages and data structures. Initially, it was just some simple Python scripts, but gradually the focus shifted to full-fledged applications based on object-oriented programming principles.
Mind you, a lot of the coding had to be done by other people because I was (and still am) a fairly mediocre developer. But even at my level, I find that moving away from Excel and leveraging more powerful languages like R has had a profound impact on the speed at which I can produce actionable insights, the depth of those insights and associated visualizations, and last but certainly not least, the way that I view and approach marketing and business problems. It’s been a profound and fundamental shift for me.
Granted, I still use Excel at work because that is still the standard tool for playing with data and sharing it with colleagues in a corporate setting. And I do occasionally lean on Google Spreadsheet at home for quick and dirty analysis (and for playing around with the Google Analytics API). But the bottom line is that I’ve more or less abandoned Excel and spreadsheets in general, so that I can build up my software engineering and data science chops.
Interestingly, I’ve had some people ask me why in the world I would invest so much time and effort into building entirely new skill sets like coding, applied math, and applied statistics when I could easily delegate these tasks to others. The answer to that question is really quite simple. I’m constantly meeting extremely young interns and new hires that have these skills (some have earned Master’s Degrees or even PhD’s in data science related fields) and these young kids will one day be vying for the very jobs that I (and you, the reader) want and/or currently have.
And I’m not interested in becoming obsolete.
P.S. If you think that you’re “just not a programmer” or “just not a math person” stop fooling yourself. There’s no such thing. I’m an English Literature major for goodness’ sake!
P.P.S. Thanks to Chris Le for inspiring me to write this post.