This month, Panoply is running an Instagram Data Challenge! The gist is, you can pull your own Instagram data into a Panoply database with a few clicks, make a visualization or tell a data story with the data about your own posts, and enter to win one of several $1000 awards.
To kick off the challenge, I did a bit of exploration into my Instagram story maker feed over time. Since I can’t win the competition… I kept my entry pretty low key.
Starting with a Common Question
I started with a common question/concern for IG influencers and brands everywhere:
“How many followers am I engaging over time?”
I made the assumption that unique commenters is a good proxy for engagement (unique post likers would have worked well too but unfortunately that’s not available from the Instagram API). I also assumed this would be a factor of the volume of posts over a given time period so to start, I ran a query to give me the number of posts I posted, and number of uniquecommenters, per month since joining IG.
SELECT "month of year", count(DISTINCT c.commenter) commenters, count(DISTINCT c.media_id) posts FROM ( SELECT date_trunc('month',m.created_time) "month of year", from_username commenter, media_id FROM instagram_comments c JOIN instagram_media m ON c.media_id=m.id ) c
GROUP BY "month of year";
I used Metabase to visualize my query. (because, Open Source)
What I found was that unique commenters is more a factor of post volume, and the emotional response from my friends and family than the number of followers I had at the time.
Typical data analysis… one question leads to a million more…
But Why?? // Time for a Story
Sometimes it’s best to dig deeper into the data and statistics to understand the relationships in the data, but sometimes a story is the answer.
Across my timeline, there are two pronounced peaks in terms of commenter engagement. The first peak in July 2016 precipitated the second in February 2017.
I asked my wife to marry me on the top of Mt Rinjani in May of 2016 but oddly enough, that posts didn’t receive a single comment… maybe because the subject wasn’t clear…
Then, a couple months later in July 2016, while back in Portland, OR for a visit, I posted about our recent engagement. (below)
The post got whopping 6 comments! A couple other posts that month helped bump up the unique commenters count for the month.
After getting engaged, Lisa and I decided it was time to move back home from Hong Kong and thus began a bittersweet time wrapping things up in Hong Kong. I played my last few playoff games with a great group of guys in the HKSPCA took Mr. @rhinoismyhomeboy on our last few walks in around Sheung Wan. Finally, I drove Rhino to the airport for the flight home (you can tell how excited he was). The caption was:
“Today Rhino begins the long trip to the green grass of Central Pennsylvania. We catch up with him again before May to bring him back to his childhood home in Portland, Oregon. He says thanks for all the love in Hong Kong#rhinopug”
There is a lot more to the story here. A lot can be found in the data and a lot can be found in the stories too. A picture can tell a thousand words, find out what your picture’s metadata can tell you!