I’ve decided to give this new fad known as ‘blogging’ a whirl – I’m not sure but I think the ‘b’ might be silent? I haven’t had much time for yogging lately so mabye I can stick at this. Joking aside, I’m pretty late to the game in the world of Data Viz but I’ve been meaning to start for a while now so I’m finally biting the bullet.
I’ve been participating in Makeover Monday for a number of months now. If you haven’t heard about it and you’re a Tableau fan/user I would strongly suggest you check it out as I’ve found it invaluable from a development perspective. Every week a new data set is published and people re-create the visualisation and share on Twitter using the hashtag #MakoverMonday.
The last few weeks I’ve been letting my new job and toddler count as an excuse for not joining in but I’m going to try to improve my participation. I recently became a Tableau featured author which shocked and delighted me in equal measure, so this has given me a renewed motivation. When I see the masterpieces produced by the likes of Pooja Gandhi, Adam Crahan and Chantilly Jaggernauth (you’ll find their work here here) I occasionally feel I should just throw in the towel. To compare my dashboards to theirs would be similar to someone drawing a matchstick man with their foot and comparing it to, say, a portrait by Monet. So I try to just use these as inspiration and glean as much knowledge from their approaches as I can. I’ve picked up some tips and tricks by downloading workbooks of others in the community (Rody Zakovich and Lindsey Poulter are two of my favourites). I’ve also found the blog posts and YouTube videos from Andy Kriebel a great resource from a technical perspective.
I don’t always stick to the rules but try to keep the spirit of the project in mind. For me, it’s about developing my Tableau skills using a data-set which has already been prepared for us. Many thanks to Andy Kriebel and Eva Murray for doing this thankless task week on week. Data preparation can often be a time-consuming process so it’s great to be able to get stuck into the analysis in Tableau (aka the fun part). While I’ve done some quick vizzes in half an hour to an hour, I often end up spending 2-3 hours when I participate if I think I will benefit from it. My primary goal is to improve my own Tableau skills rather than focusing on trying to improve on the original.
This week the data-set came from a table on the Social Blade website and I have copied an extract of what is presented below:
3 things I liked:
- The list is ranked and allows flexibility by giving people the option to sort on different values
- The information was very detailed with data for the top 500 users
- If you clicked into a user you could see a time series graph and lots of other KPIs
3 things I didn’t like:
- The view was essentially a wall of figures which tends not to be very engaging or provide insights
- The column headings weren’t aligned which I found a bit confusing initially
- When I sorted on the SB scored I found the numbers weren’t consistent. There were a couple of very high SB scores which could be anomalies/outliers – in a professional setting this would have required further follow-up but for the purposes of this exercise I don’t vet the data, I simply focus on the visualisation
3 things I’ve tried to do with my viz:
- I wanted to use the cool new feature in Tableau 10.2 which allows multiple colour legends for measure values – this gives life to tables, allowing for comparison within the column and showing trends across multiple measures
- When I first saw the data I thought a scatter plot would be interesting to see as it’s great for showing relationships between multiple measures. I also thought a treemap might give context. I created both of these but left them out of my final dashboard as I didn’t feel they fit with the design this week. There was some discussion on Twitter around treemaps last week and Jeffry A. Schafer wrote a great write-up on these if you want to read more about their use
- I thought summary information would be useful to allow people to get a handle on the data-set contained in the table. To achieve this I’ve provided max/min/avg figures and created histograms
I’ve noticed a lot of people going for a style which is suitable for mobile recently so I decided to trial that format this week. The result can be viewed here in Tableau Public and also can be seen below:
As Andy Kriebel pointed out recently in a blog it’s good to think about who your audience is. In the case of this website, it could be the general public interested in the ratings or it could be a specific user interested in their results. This viz is designed as a general overview – the user would get more insight from going to the website.
That’s it from me! Comments or feedback are welcome 🙂