The Numbers are Not the Whole Story
Ask anyone interested in politics or sports—there is no shortage of data available. Of course, the same is true of companies like Amazon and Walmart that process millions of transactions a day, as well as social media sites that feature countless posts, photos, likes and comments. One of Kenway Consulting’s key capabilities is Information Insight. Through data governance, management and business intelligence, it is our goal to change data into information that is easily consumable by our clients.
Before I get into my personal journey of turning data into information, I want to take a look back at a recent event where the data presented was misunderstood—and continues to be misunderstood today.
Donald Trump’s victory on November 8 was an upset, but it was never said to be impossible. The odds makers, on average, predicted a Clinton victory at 85%. She lost, and now, by and large, political pundits are wondering how everyone was so wrong. The odds makers were not wrong; rather, the improbable happened. Trump won every historically republican state, every “battleground” state except for New Hampshire and a few that had historically voted for the democratic candidate. That is not to mention that at the time of writing this blog, Clinton has more of the popular vote overall than Trump. People did not believe that the 15% chance of Donald Trump winning would happen. However, it seems misguided for the same person who buys a lottery ticket with a one in 175 million chance of winning to also say that a three in twenty (15%) possibility occurring is impossible.
I bring up this case to show that people often do not understand little data. How can we expect to understand big data without first understanding little data? The key is to not only put data into a consumable format, but to also fully understand what it means.
To that end, I have completed four century (100 mile) bike rides, and I have data on every mile of every ride thanks to my Garmin watch. The rides were done once per year between 2011 and 2014. I can sync the data into a spreadsheet, open it up, look at it and learn absolutely nothing from it. At this point, the data is simply numbers. I plotted my overall time to 100 miles, which led me to believe that my performance improved from 2011 to 2012 greatly, and then I maintained that improvement the next two years. To test this theory, I used Qlik, a leading business intelligence tool, to put the data in a more consumable format for me to analyze.
The important question is “am I getting in better shape?” I can see from my overall moving time that each ride has been faster than the last, so I’ve ridden faster, but what does that mean for my overall physical fitness? If I assume my moving pace to be a measure of my performance, and my average heart rate as my effort to achieve this performance, I can quickly assess the performance vs. effort required. I created graphs of the two measurements, put them side by side and observed the following:
My biggest improvement in pace occurred between rides 2011 and 2012, but during that time, my average heart rate also increased from 140 to 148. From 2012 to 2013; however, my pace improved slightly, but my average heart rate dropped back to where I was in 2011. Between 2013 and 2014, both my pace improved and my heart rate increased slightly. This would suggest that I improved my fitness level between 2012 and 2013, and I maintained that improvement in 2014.
Another test of fitness is the number of calories I burned in each ride. Where my average heart rate tells me how hard I had to work, I see calories burned as a measure of how much I can work. Therefore, a higher number is better in this case. To demonstrate this, I created a pie chart showing percentage of total calories burned for the four years separated by percentage in the chart. In 2011, I burned 22.3% of the total. This increased to 25.4% in 2012 and to the highest at 26.3% in 2013. 2014 again had a slight decline to 26.1%. This would suggest that in 2011 I was in the worst shape and in 2013 my best.
If I combine the two analyses, I can confidently say that I was in better shape in 2014 than I was in 2011. The largest improvement occurred between 2012 and 2013, with 2014 maintaining that improvement.
This is a short, but useful, example of how the numbers alone do not tell the whole story. Simply observing the time from my rides would have led me to believe that I made my largest fitness improvement in 2012. However, as I view multiple metrics, I can see that I worked a lot harder to achieve the improvement in 2012, and it was in 2013 where I was not only able to increase my speed, but also drop my heart rate back to 2011 levels where I had started. Whether it is little data or big data, it is important to understand the meaning of our measurements. By doing so, we can turn big data into consumable little data, and then use real-life examples to understand that little data to make better informed decisions.
We love talking data and analytics at Kenway Consulting. If you would like to continue the conversation on data analysis and strategy creation, or any other BI or Analytics platform, let’s get in touch: firstname.lastname@example.org.