Last year I wrote a blogpost describing Excel as the Dennis Rodman of business intelligence tools. In the blogpost I described events where data management and data scientists got together, and Excel skill was welcomed in a very odd way. Like the NBA player, Hall of Fame inductee, 5x champion, Dennis Rodman, Excel wasn’t taken seriously, was seen as a bit of a troublemaker. Nevertheless, it’s hard to achieve any meaningful victory without Excel even if its only role is to hold data. (But it is doing so much more.)
Rob Collie of PowerPivotPro has joked:
Question: After “OK”, what’s the second most used button on any BI tool?
Answer: Export to Excel
Yet, there’s no shortage of Excel haters. Like spurned lovers, they sit and generate lists of slightly true, extremely contrived and purely invented reasons why Excel is horrible.
But there’s hope!
PEACE IN THE PDX
Since moving to Portland, Oregon in August 2014 my experience has been quite splendid.
Portia even suggested the book Data Smart by John W. Foreman. It’s a data science book that’s 100% Excel and very little VBA coding.
Bottom line: it’s a beautiful thing to have sober discussions that aren’t sprinkled with insistence that Excel is the devil.
EXCEL vs R
One of the data scientists at this weekend’s meetup offered a comparison of R and Excel.
The complaint he had about Excel was the potential that formulas and data can be in hidden rows, hidden columns, hidden sheets and named ranges. With R, the code is all in one place. The downside with R is learning how to write and read the code. Excel formulas are so much easier to read.
Is one better than the other? That’s actually a silly question. There are just the realities of the respective tools in service of tasks. What I see are grand possibilities when we have discussions like this. There’s opportunity to work together and make each other’s lives easier. If I was sending spreadsheets to him to process in R, I’d make it a point to eliminate or minimize anything hidden or scattered.
AGAIN: PEOPLE, PROCESSES AND TOOLS
This is related to a previous blogpost, Data Management in Perspective. It was all about understanding the interplay of people, processes and tools. When 2 people can agree that there won’t be any hidden rows during our hand-off, this particular process can keep moving forward while accommodating our respective BI tools.
We have choices. We can tell lies, bicker, feud and blame or, we can realize that we’re all on the same team as Excel developers, R coders, database architects, etc. There’s too much at stake because data is everything. It’s household budgets; it’s forecasting how many nurses you should have on a shift; it’s determination of whether you can open and sustain a new office; it’s inventory of who gets what in a divorce; it’s an airline’s forecast of fuel costs. The tool used for calculation is only as good as the people and the overarching processes.
I’ve met other good data scientists here in the PDX that give me hope that we just want to win games with whatever we show up with. Like Michael Jordan, Scottie Pippen, and Dennis Rodman, we can do this!