Using Tableau to measure the effectiveness of my New Year's Resolutions
This was one of my New Year’s Resolutions for 2018:
Pay for less coffee.
Now please note that I didn’t say drink less coffee, just not pay for it. 😊 (And no, that doesn’t mean I’m robbing banks to fund my caffeine intake.) It means simply bringing your own coffee from home (much cheaper per cup) or taking advantage of the free coffee available at work, at client offices, or during networking opportunities.
I feel like I did a good job this year.
I know I had at least a one month stretch where I was making cold brew at home and bringing it to work. I also had clients with free coffee (or very inexpensive coffee) which helped me spend less. I also went through an Earl-Grey tea kick over the summer, so that should have helped too.
But how did I actually do? For that, let’s look at some numbers.
Pulling data into Tableau (more on the data prep below), I wanted to quickly compare 2017 to 2018 and look at total coffee spend by week and use a running total so I could see my total spending over time. The results are below:
While it appears 2018 has slightly less spending at Week 48, the overall chart looks pretty much the same. The vertical gap between 2018 (pink) and 2018 (purple) lines represents the difference in spending to date.
I’ve spent $116.95 less on coffee this year but the rate at which I am spending has not changed.
So, the conclusion? Gordon has managed to save $116 this year because of his New Year’s Resolution. So while he feels like he did a good job and has spent significantly less, the numbers just don’t tell the same story.
Being Honest: This is exactly the sorts of things I hear from clients all the time, but about their data and business performance. "I feel like we are doing so well...." but their data tells another story.
So what's the takeaway?
Without numbers, you base performance on feeling and perception. While "Gut Feel" is great in some scenarios, research shows the lack of immediate feedback, lack of familiarity, and lack of practice will lead to poor decision making and faulty conclusions in your specific scenario.
So, did I really achieve my New Year's resolution?
Yes, about $116 dollars of it, but not the $1,000+ dollars I had envisioned in my mind.🤑
Find Gordon elsewhere online:
Tableau Public: https://public.tableau.com/profile/gordon.strodel#!/
Here’s how I prepared the data:
Download transaction data from my Mint.com account (I downloaded 4 different categories and descriptions in 4 different files…long story.)
Merge Files and filter out any obvious non-coffee vendors, example: Qdoba
Aggregate the data (group by all fields) to de-dup the transactions which might show up (my mint category files were overlapping)
Exclude any transactions with amounts over $20. I found these were lunches, coffee goup-bys for work, coffee purchases online (whole bean) or other random stuff
Group the vendor names together to simplify analysis (combine “PEET’S #05305” and “PEET’S #29205” into “Peet’s Coffee”)
The final result was a CSV file with data from 2011-now and 1,083 transactions!
Looking at the Top 10 Merchants over time as it shows some interesting trends:
When I joined Slalom in Jan 2018, I was suddenly near a Starbucks, Flour, and Davio’s, so you can start to see me spending more at those vendors.
My clients in 2018 influenced the trends. Fresenius had its own café. And the really recent uptick for Boston Common Coffee is when I was at Rapid7 and BoCoCo had a café right around the corner on Washington Street.
When my prior employer moved offices to North Station, they were right by a Boston Common Coffee and an Equal Exchange but not a Starbucks. My BoCoCo spending went up (see that green line!) and Starbucks slowed down (but was not eliminated due to my client travels.)
You can look back at when my kids were born in Summer of 2012 and see the rapid rise in coffee consumption at Starbucks!