Building accessibility into a dashboard is something that is hugely important but rarely done properly. This week was my first attempt to build accessibility into a dashboard from the ground up and, whilst a lot went well, a lot could have gone better.

My philosophy was to incorporate accessibility from the beginning, with filters and keyboard/screen reader accessible elements at the top of the dash and charts at the bottom. This actually went fairly well, my main issue was that I bit off more than I could chew with the concept of my dashboard. The fact that I spent around a third of the day building the concept and joining relevant datasets, and then another third building and messing with calculations meant that I had relatively little time to ensure that the accessibility features were up to scratch. This meant that during the presentation several of the features did not work, such as the screen reader readable table. Similarly, the lack of time meant that I had relatively little time to ensure that all of the filters were all working exactly as intended, and didn't have enough time to fill in some of the captions that I wanted to to better describe the viz.

Meanwhile, the retrofitting was something on Monday was something that was far easier. As I wasn't working against myself I was able to focus entirely on the accessibility, making sure that it all worked to the standard expected.