A visualisation of Kirsten Alana’s Instagram data with the aim of providing insights to help improve engagement. There are two separate dashboards, one for specific insights and one for exploratory analysis. They are essentially the same but the specific insights dashboard points out at least one insight per chart whereas the exploratory dashboard has additional filter options and allows you to find your own insights.
The analysis is broken down into 3 metrics, likes, number of comments and comment sentiment. For likes and number of comments the avg per post is mostly used whereas comment sentiment gives an avg polarity score from -1 (negative comments) to 1 (positive comments). Each metric has its own section but these are similar across the three. The first chart answers where is best to post, the next when (time of day), and then finally how (filters/tags?). Additionally each metric has a bonus chart for some deeper analysis/exploration of data.
Hovering over any chart will display tooltips with further information. Find the interactive viz here (full screen option at the bottom of the page).
In order to prep the data I didn’t use SQL. I instead used the ETL tool Alteryx and was impressed with how easy it was to connect with Panoply as well as how fast the data ran through the tool. For anyone familiar with Alteryx I can post a screenshot of my workflow (was going to post it here but I’m limited to one image as a new user).