This blog is part of the Dashboard Week series. If you want to know more about what the Dashboard Week is, have a look at the first blog in the series.

 

Coming off a high of producing an explanatory dashboard that I was satisfied with for Challenge #02, I was kind of looking forward to our third challenge.

The Challenge

NSW Public Library Services produce a rather comprehensive set of statistics every year, and upload them on their website for public access. We were challenged to sift through the pile of goodies and find something that we like, and then produce a Tableau dashboard to visualise the information.

The Data

One of the main obstacles we faced was the structure of the data. Every year’s data is in one Excel file, with each metric in a table on one sheet, and with some metrics divided across a few sheets. The primary level of granularity appears to be at the council (LGA) level, with each of them running one or more libraries, although such details are not available in the spreadsheet.

To further complicate things, the councils did not all match up across all sheets from a single year. I decided early on that, (partly) in the interest of time, I would build my dashboard on a cross-sectional slice (i.e., not longitudinal) of the data. It turned out to be an inspired decision, because I could spend more time on honing my design skills. But more on that later.

I was still undecided regarding the metrics to focus on but I knew that I wanted spatial data, so I went and scraped the locations of all the libraries listed on the NSW.net website.

After (finally) identifying the metrics I wanted to work with – which are mainly related to the service aspects such as information requests and number of visits – I picked through the Excel file, moved the sheets onto a new file, and then stripped away the unnecessary rows and columns.

And off I went to start building my dashboard!

Visualisation

Regular readers (of whom I have none, I’m keenly aware) would know that I always have some sort of design philosophy when I’m going into a build. Having just completed a traditional dashboard befitting the theme of the data on Day 02, I really wanted to change things up. I have been wanting to create something that is reaaally un-dashboard-like for a while now, so I thought, “hey, why not do something crazy during Dashboard Week!”

The broad idea was to have some sort of a starting dashboard (a homepage, if you will), with no charts and minimal text, and yet it should be clear to anyone what they should/could do next. Initially, I thought of making something that looks like the Google homepage. You know, the universally-known search bar with a couple of words typed in and a bunch of auto-complete options in the dropdown. Most people would guess that they should try clicking on the said options.

However, in a moment of inspiration after seeing what my friend Julián initially picked for his dashboard background – a black-and-white sketch of bookshelves full of books, I thought I could get something similar but with a few different elements in there (e.g., a librarian’s desk or things hanging on the wall), and then fill a few elements in with bright colours so that they would stand out and scream for attention.

You would probably have guessed by now that data analysis was not on the top of my priority for this project, so it ended up being an exploratory dashboard. But the most important thing for me is to try something new and get some experience under my belt, and I am pretty happy with how it turned out. The end result is a homepage that looked like this:

 

 

J Tay
Author: J Tay