0

Data can be fun and engaging, don’t laugh! It can spill out of your analysts cave, and help you optimize not only your business but also your enjoyment. Yes I just said “optimize enjoyment”, like enjoyment is a mathematical problem that has an optimal solution. So the ruse is up, you’ve caught me red handed, I’m guilty of being a data nerd. I openly confess to using data in my personal life to help make decisions, and I’m willing to bet that you do too. The trick for analysts, like myself, is to tame ugly wild data and dress it up all fancy. This is how we fool the non data nerds into unknowingly using data to help decide what to do or where to eat.

As an example of data improving personal experiences and helping inform decision making we need look no further than Cincinnati’s own MidPoint Music Festival (MPMF). Not even your most informed, music snob, hipster friend is familiar with all the indie bands that perform at MPMF. I’m enthralled with experiencing new culture, but with multiple stages and a lineup including such bands as “Grizzly Bear” and “Machines Are People Too”, I didn’t know where to begin. Suddenly I’m paralyzed. What if I watch band X but all the good people watching opportunities are happening with the crowd at band Z’s performance?

This is the time when my data nerd self springs into action. This isn’t my first rodeo. I’ve followed events twitter hash tags, witnessed projectors streaming a live feed of tweets & facebook comments, and gotten the scoop on what friends are doing via their check-ins. The people at these events are creating a real time feed of data that gives streaming insights about what is going on and what they think about it, so the opportunity is sitting in my lap. I can tap the data from their mobile devices and social media posts, but my challenge is to properly collect, aggregate and organize this data to make sense of it on the fly.

Enter Tableau, a nice analytics and data visualization tool that plays well with data regardless of its size and format.  I had my platform, now I just needed the data. I decided to leverage the collective ‘expertise’ of the attendees and use twitter activity as an indicator of crowd-swell. Twitter is an unstructured data source, which can be a formidable obstacle, but I had chosen my weapon wisely and was up for the challenge of turning it into something useful.

Given that I knew all the band names, stage locations, and set times my approach was as follows:

  • Capture tweets containing @MidPointMusic and #MPMF (In a perfect world I would have collected all tweets including the band name hash tags as well)
  • Parse the text of the tweet for any band name, when a match is found count that as a ‘mention’ for the band
  • Overlay the mention data on a map of the area with the stage locations. In theory the result creates a living map that calls attention to the locations with the most stuff going on.

Click through the date/time filter on the embedded visualization and you can witness what bands and locations garnered the most buzz over the course of the 3 day event.

[iframeWrapper url=http://public.tableausoftware.com/views/MidPoint/Dashboard1?:embed=y&:display_count=yes height=800 width=639]

So here’s the breakdown of what just happened:

  • I needed help making decisions
  • People are spewing data into the cloud that is useful for decision making
  • Harnessed the data and cleaned the spew off
  • Wrestled it into an informative shape
  • Produced an output so easy to understand a caveman could do it

If you’re a caveman or a data nerd like me – and want more info about my collection methods – leave a comment.  And if you watched a band and were part of that crowd-swell at MPMF, tell me about it.

Leave a Reply