Screen Shot 2020-08-29 at 7.28.28 PM.png
 

 Social App

iOS application created for Apple teams
to quickly get feedback from Social media about their projects.

 

Time-frame: 10 months ( Aug 2016 - June 2017)

Team: Applied Machine Learning team

My role: User experience design + Visual design

Deliverables:

  • Exploratory research

  • Story mapping

  • Interaction design and visual design based on Human Interface Guidelines

  • Testing with users

  • Iconography

Problem

 

Everyday people leave feedbacks about the products on social platforms like Facebook or Twitter. These feedbacks can be very useful for the team that works on this product to improve it, but it’s impossible for human to monitor all posts on social media.

 

Web Dashboard for business experts.

With machine learning technology Social project enables social media listening and monitoring. It consists of Dashboard (web platform for viewing and analyzing social media) for business analysts and iOS app for sending alerts to business teams managers.

I needed to design iOS app from scratch.

iOS app for 👩‍💻👨‍💻 Apple teams and managers

The original web application was designed for business experts, and the amount of information, provided through web interface was excessing the needs of Business Team Managers, who was my target audience. 

The challenge was to present information, in a way, so people, who didn’t have experience in analyzing social media, could easily act on the provided information.

Case_management_social.png
 

Research

I started from learning how this app will fit in business teams life. What exactly managers want to get from Social Media.

I conducted interviews with the teams and teams’ managers. I showed them web dashboard and tried to understand what information teams could use.

User pain points

  • Business teams don’t have time to learn new tool;

  • Teams are interested in real time information, but they can’t spend a lot of time at Social Media space to find it

  • A lot of people are not familiar with social media platforms, but they just want to see feedback from users.

 

Use cases

  • New Product Release. Teams need to see all problem around the event, how start of sales happened, did people find out something frustrated

  • See feedbacks about resellers stores

  • Control informational leaks

  • Get positive feedbacks to support team’s mood

I talked with potential users and asked what use cases they see. One of the examples was: “iPhones are sold out in AppleStore, somebody posted about this on Twitter, we want to resolve this problem quickly by sending this issue to retail manager who is responsible for this.”

Teams that were going to use this application were not going to analyze trends, they needed alert/notification about particular problem to solve it quickly

As this type of use cases was most common, we decided to use it as a starting point for design.

 

First users flow diagrams

IMG_1240.jpg
IMG_7548.jpg
 

Users needs

  • User should be able to read about the app and download it

  • Users should be able define main keywords and metrics to receive actionable information

  • Users should receive real time alerts about events at social media

  • Explore details about this problem - there are details like sentiments, number of people talking about this problem, volume trend, triggers to make a decision

  • Links to posts at Social Media - user can open links to posts or comments at twitter, reddit and etc.

  • Leave comments for teammates - sometimes user want to add some details or comments to their teammates, while they investigate problem

  • Share the problem - user may want to share information with teammate responsible for this or share it with manager

  • Save information - user may want to save problem and to investigate it later

  • User should see what problems are new

  • User can delete any irrelevant alert

Machine Learning team point of view

  • To improve model we need to control feedback from users, we had to track what alerts did user open and escalated to a team

Teams flow

With product manager we defined how users would be able to escalate an issue, in case it’s important and should be fixed ASAP.

 
user_story_updated.png

Wireframes

Machine learning team could get a lot of information about each post; I decided to create two sets of wireframes. One option would show more details about each post and the other one with minimum information.

Option 1. Show post with basic details about it. Description, Triggers and Post

 

Option 2. More information and details about problem. Sentiments, trending, triggers, top tweets, retweet.

 
paper.jpg
wireframes2.png

I put wireframes into simple prototype and showed to our potential users.

A lot of information about alerts were useless for them, they couldn’t understand what all this numbers mean and they didn’t want to learn one more new tool. So we started to implement pilot version with basic information for alert.

 

Pilot Version

  • Alerts are organized as a list, latest are on top; this way it’s easy to scan titles

  • User can save/bookmark alert to see dynamic of this post later

  • Triggers show why system thought alert is important

  • User can escalate alert to predefined list of people

  • Escalated alerts marked as escalated and has history of escalation 


Pilot version we released 👉

Based on internal app for tracking user activity, I could see that users almost always want to see all posts attached to the alert.

pilot_small_frame.gif
 
 

Updates

I updated design with some changes

  • List of posts was moved to details screen. It became easier for users to read them without extra steps

  • Moved “escalate” feature out of share flow to improve escalation flow for users; this also helped us to highlight this feature to users.

uploaded_v3_textovaya+4-1.gif
 

INTRODUCING CATEGORIES FOR ALERTS

As more teams started to use the app, our team continued improving the app.

  • Adding new “closed” status for alerts; each escalated alert can be closed after ticket was opened or if alert was irrelevant

  • Categories help users to organize alerts in a better way and control how many were escalated, but still not processed

  • Minor visual updates

2018.png
 

Results

In 2018 app was already used by 12 teams. It helped to catch some critical leaks about new apple products and quickly stop spreading them out.

19% of sent alerts were converted to tickets.

What’s next

As number of teams who wants to use it is growing it is important to create system, so teams could onboard themselves to start using the app. It is also important to give opportunity to teams to define keywords and topics they want to control, but don’t create complexity in the app.

 
 

Add on: App Icon design

Icon should catch users attention and reflect purpose of the app. Here are the options that I created in the process.

logo_social-03.jpg