Hello
I am Gonçalo and I use Instagram. I don’t post everyday, but when I do, I don’t do it at 4AM. Why? Because at that time, many people will miss my photos! And it’s all about sharing, right?
Having this in mind, my mate Rui Magalhães, who is a photography enthusiast, constantly sharing his work, came up with an idea of having an app that analyses his posts and tells him the best options to make it successful. Wouldn’t it be nice? (:music:)
Whitesmith Hackathon
Let’s rewind a bit: 3 months ago Whitesmith was turning 4 years old (:tada:) and to celebrate that milestone, we decided to have a full week retreat, where we would be working in a remote valley (with Internet). By the end of that week, we held a Hackathon, and that’s when Rui proposed that idea. Seeing the purpose in that, I decided to jump in and help with what I could.
The idea was simple. We would use Instagram’s API to correlate the success of a post with its meta-information: tags, hour of the day, location, filters. To calculate the success of each post, we created a unit, Qarma Points, which is very simple to understand: each like earns you 1 point, each comment earns you 2. This is because liking a post is much easier/detached than commenting. So, even if a comment is not favourable, it surely means someone engaged in that post for some reason.
How
To do this, we decided to mix a bit of our own skill-set with trying some new things. João Barbosa joined us with his Rails super-powers and built an app from scratch, while Rui was experimenting with design (using Bootstrap, Javascript and CSS) and I was doing a bit of both, helping João at the back and Rui with a HighCharts chart to display the results of our analysis.
Since Instagram’s developer API only allows 10 users and the 20 most recents posts to be fetched, we were quite limited for testing, but it did the work. With the Qarma Points logic I explained above, here’s the app finally working:
Yay, now I can analyse my metadata and see which posts where more successful! This is fun!
Results
We presented this to our peers and everyone was excited with the ability to try Qarma with their own Instagram account but… to do so, we first had to submit Qarma for Instagram approval, and that’s where the fun ends. Despite our efforts to comply with what was asked of them (Privacy Policy, Terms of Service, a video showing the app working, a description of our purpose using the Instagram API..) we we’re rejected because apparently, they couldn’t fully understand the use case (I know, right? :/ ).
Until now we didn’t submit again and the project lost a bit of traction. We’re still deciding whether we should kill this or give it another try. Do you have any thoughts on this?
Nevertheless, this was an amazing experience. For me, being my first hackathon and as someone that enjoys both doing some meta work and also programming a bit (although I am not skilled as Rui or João :star_emoji:), it was great spending two days with my friends, building something from scratch.
This article is the fourth in a series of posts about the first Whitesmith Hackathon. The other three are Hawkpost, Quoi, Aquda.me.
The Non-Technical Founders survival guide
How to spot, avoid & recover from 7 start-up scuppering traps.
The Non-Technical Founders survival guide: How to spot, avoid & recover from 7 start-up scuppering traps.