Last summer, when I started to formulate the idea for Real World Fake Data, I wasn’t quite sure how it would be received, what the participation level would be like, and what the end result would be.
The original inspiration came from a request that I had received as someone was looking for an example of an industry dashboard. I found was a bit lackluster, and honestly, there was not a lot of what I would think to be business examples of dashboards in tableau public. And when you start to think about this it makes a lot of sense, because business dashboards are generally driven by business data. Business data is sensitive, therefore, it can’t be shared out into the wild. But, Tableau Public is always the first place I go to find examples of dashboards, so I started to think about how I could encourage the community to create business Dashboards?
I realize that the biggest hurdle for this idea was data, and luckily I discovered Mockaroo and I started to experiment with creating fake data sets. Generally, Mockaroo is used for creating data sets for testing web application forms and things of that nature, at least from what I can tell.
Mockaroo quickly became the seed of what would become real-world fake data. I started to make a list of different industries that I wanted to highlight, and industries that I could find examples of for creating these data sets. Once I came up with my list I ended up with 12 potential data sets, now I was able to find a couple in the wild that were generally shared like the financial consumer complaint stayed up. The others I was on my own and had to create them.
So I researched and came up with a very simplistic fieldset, realizing that I would never be able to truly mimic a real scenario data set, because quite frankly business data is specific to that general business and that particular company. But doing a complete mimic of a business dashboard was not what I was after, but I wanted to have templates that we could use his aspiration. in the end, I think we got that.
Once I figured out what the cadence would be as far as releasing the data sets, I started to tease and advertise it a bit to see if people would be interested. Right away, I was happily surprised to see support from the data community and excitement around what the project could be.
With everything in place, the data sets complete, and me being ready to embark on this six-month journey I released the first day. I was so happy to see not only the level of participation but participation from the community members that I haven’t seen before, so it started exposing me to new visualization authors which I thought was amazing.
Additionally, the community that was participating did so much in advertising the project as well, with every LinkedIn post and every tweet they would always share the link to the information page in the data said and it was one of the things that just made it work.
There were undoubtedly a couple of dataset duds! The supply chain data set, was by far the least visualize data set. With that said, it was also the most complex with having basically an entire star schema within the Excel file, so I can’t say that I am shocked, but I did take it as a lesson learned. certain industries and departments obviously had more interest in participation like finance and insurance and human resources, and then data sets like solar energy and hospitality and the emergency room data set while more niche in nature still got really great participation.
Here are the lessons learned from launching, and executing a community project.
- Try and model more realistic data for the data sets. While the data sets were randomly generated, in most cases a lot of the data looked flat, and to make them more interesting having a little bit more variability would’ve been nice.
- Don’t go to niche! While I was really excited about some of those “oddball” data sets, people didn’t dive into them the way I hope they would, and that’s OK.
- Bring a friend! While it was cool to run it completely by myself, a couple of people had suggested along the way that I get a partner in crime, but for simplicity’s sake, I decided to take it on my own. And it was fine, but I think I would’ve had more fun if I had a partner so maybe next time I’ll do that.
And that’s it, it was a lot of fun and I’m glad I did it. It was a great experience running a community project, and if you’ve got an idea for one just do it!
I do wanna thank again everyone who participated because without your participation this project would have fallen flat and it would’ve died very quickly. Yeah, we made some amazing dashboards, and I know that they are used and referenced by a lot of people in the community, and a lot of people inside Tableau. I’ve heard from folks in the training department, in customer success, and in the solutions engineering pre-sales team, and they all love the project, and have been able to utilize some of your examples for teaching and inspiring others!
And I want to thank everybody for all the feedback they gave on how much they learned and how much they enjoyed the project, Emmy and that’s what really made it worthwhile for me knowing that I helped either you directly in your learning journey or someone else as they’re trying to find inspiration for building their business Dashboards.
People have asked, “will there be more Real World Fake Data?“
At this point, I don’t know. But, it’s on my radar and in the back of my mind. And if it does happen you’ll find out about it here first!