Michelle Lin

Concept: Instagram

Creating an Interactive Analytics

Dashboard for Instagram


Project Overview


We all know Instagram. We all use Instagram. We all love Instagram. What started off as a simple photo-sharing social media platform is now a digital product that allows thousands, perhaps even millions, of users to profit through product/company sponsorship.

Being an eager learner, I decided to challenge myself to design an in-app, single-screen analytics dashboard for Instagram which would allow users, both those who are incredibly active and profit from sharing photos (I think they're called "Influencers," these days?), as well as everyday active users, to track the various trends which mattered to them most.

Using my network, I was able to connect with 10 "Influencers" who had thousands of followers on Instagram. Since many of them did not live in San Francisco, I created a Google Form survey to better understand these users — from there, I taught myself Flinto and was able to achieve my learning personal goal. 

One-woman Product Designer!




As an frequent 'grammer, I wanted to created an in-app analytics dashboard which retained the same UI style for the photo-sharing app — one targeting those with or hoping to get thousands of Instagram users. I conducted 10 user interviews and found the features which mattered to these particular users (or so-called Influencers). Based on the user research findings, I decided on using smooth line graphs without points for the data visualization as a means to fit the new and simple Instagram UI. It also gave me a chance to continue mastering the art of creating Bézier curves. Using Flinto, I was able to create an interactive prototype that incorporated horizontal scrolling and in addition, I added a popover to explain the visualization. 


Design Process




I spent the first portion of the project researching dashboards of social media apps (i.e., Twitter, Medium and even Iconosquare), but struggled to find a mobile dashboard for applications of the sort. And no, I do not own a Fitbit. In any case, I then moved on to developing a persona based on assumptions and real-life Influencers. Moving forward, I conducted user research to better understand what metrics mattered to those who fit that persona.

Researching Dashboards

Beyond simply looking at the dashboards of similar social media applications (all of which, that I know of, were online), I also ventured out into the world of third-party social-media analytics websites such as Iconosquare and Snaplytics; however, I was unable to find something strictly for mobile. 

Because of Instagram's latest UI transition, I sought after the simplistic look and felt the most drawn to Twitter's dashboard which were very simplified line graphs which showed trends over a "28 day summary" — including "Tweets," "Tweet Impressions," "Profile Visits," "Mentions," "Followers" and "Tweets Linking to You."

Persona Development

I created a persona, Jenn, that was heavily based on many of the people I follow on Instagram who have achieved "social media-fame" to some extent. By developing a persona, it helped to guide the rest of the design process along by providing a foundation for understanding user needs and motivations.

User Surveys and Findings

Using Google Forms, I created a user survey with questions I felt would best help me understand the typical flow an active user goes through from the time they open the app to when they exit it. Luckily, I was able to gather this data from 10 incredibly active users — 50% of whom were paid to use the app as Influencers.

Based on the results of the surveys, I gathered the following information.

 1). The average number of times an  Influencer  opens up Instagram per day 2). The Instagram features valued most in order of importance from top to bottom 3). The percentage of those interviewed who receive endorsements versus those who don't

1). The average number of times an Influencer opens up Instagram per day
2). The Instagram features valued most in order of importance from top to bottom
3). The percentage of those interviewed who receive endorsements versus those who don't

The two venn diagrams ensured that I was targeting the right users when conducing the user research and surveys. Secondly, by knowing how important each Instagram feature was to the Influencers, collectively, it allowed me to design the dashboard with their values in mind.




This portion of the project consisted primarily of creating a very simple task flow and creating wireframes — which was rather difficult as I had to decide upon a single type of data visualization out of many in order to retain consistency. 

The Standard User Flow of an Influencer

Pen and Pencil Lo-Fi Wireframing

  Click to enlarge!

Click to enlarge!

Ultimately, I chose the fourth iteration of the wireframes because it best fit the desires of the users I interviewed by showing trends and optimization times. There's also an additional section which shows the photos with the most engagement (—most liked and commented). Furthermore, I added a popover to explain what each portion of the analytics screen referred to.




As simple as this project may seem, it took a little while to learn Flinto as a prototyping tool — in any case, I wanted the Hi-Fi prototype to include horizontal scrolling as well as the popover animation. The final prototype, which was created on Flinto and exported as an MP4, then prototyped via InVision, can be seen here




In this day and age, everything we do — every click, tab and swipe we make is being recorded. Welcome to the age of Big Data. With this in mind, as a UX + UI designer, we must not only study data visualizations and the types which best fit each application, but learn to design graphs and charts in a simplistic manner for specific users. At the end of the day, it's all relative. If you're working for a company like Iconosquare or Pixlee — almost every type of data visualization (i.e., histograms, pie charts and line graphs — even combinations of several) come into play. However, for this project — I stuck to simplicity and the factors and features which mattered the most to the users I interviewed.