It’s Tuesday already, and we are onto day 2 of Andy Kriebel’s favourite week – Dashboard Week – where he offers us up a data set set each morning for us to clean and produce a viz out of by the end of the day.
Today’s topic – pick either beer or wine.
Seeing as we are fuelled (ever so kindly) by The Information Lab on a self-replenishing supply of free beer, the choice was a ‘no-brainer’. I opted for beer and, seeing as it’s my favourite brand in the fridge, I chose to look more closely at Brewdog – the British craft beer start up that begun back in 2007.
The trick, that’s becoming ever more obvious as the week continues, with making a dashboard is to source and clean the data you want to use as quickly as possible – as this allows you maximum time to concentrate on the final output, a dashboard, within the day.
However, today – this part of the puzzle took me longer than I first envisaged. Andy gave us one or two sources that we were free to investigate, but I couldn’t get either API firing – and so, I entered a round the houses search for any relevant data that I could find. I must have manipulated at least 5 different ‘BeerData’ sources, but decided to drop each for reasons varying from having some missing years, a lack of fields, not fitting the brief I had set myself, etc. There seems to be a large amount of information on the American craft beer market, but much less on the UK equivalent.
Anyway, come 2 pm, I decided enough was enough and made the executive decision to simply pull information off Untappd’s website – a truly impressive source of user-submitted reviews regarding craft beers. The data took an hour to clean in total – I started first in Excel with some simple manipulation, then moved to Alteryx when I wanted to get down to the data polish.
Here’s my annotated Alteryx workflow; nothing complex just good old-fashioned cleansing:
Create the dashboard.
Once complete, I imported the data into Tableau to start toying with vizes.
From first thing in the morning, I had a clear idea of a structure for my dashboard – more in terms of graph output than specific dashboard structure; however, this didn’t take long to create once I had my elements complete.
It was almost organic with placement – in this case, and seemingly I do this regularly now, the main list of the subject formed the mainstay of the chart to which I anchored the other elements to. I do this to ensure the viewer has the simplest experience when interacting with my viz – as I like to keep instructions to a minimal and make them as intuitive as possible. It frustrates me no end having to view vizes that you feel need to come with a manual on how to gain insight from their workings.
Using Shape Files to insert interactive images.
A key learning point from today’s exercise is in using ‘Shape Files’ to insert images (.png files) into vizes. This comes in particularly handy with customer logos, for example, which you can add instructions to utilizing the tooltip function.
To do this:
- Open new tab in Tableau
- Insert new calculated field called ‘Image’ with the same exact text contained within
- Open your ‘Tableau Repository’ via File Explorer
- Insert logo (.png) file into folder
- Insert ‘Image’ calc field onto sheet
- Change mark type to ‘Shapes’
- Select shape
- Then you can simply drag onto your dashboard in ‘Float’ mode and place wherever needed
Before I knew it, I was done – to my surprise, as I had assumed it would take me a long while to add my elements into the dashboard. I used a mixture of fixed (to keep the charts in place) and float (with pictures and floaty bits) to create the viz.
Here’s the finished article – making full use of Brewdog’s colour scheme (that aligns rather nicely with my apparent love of the colour Cyan):