US Army Data-Driven Bot Detection
Inauthentic, bot-like activity on social media can skew metrics making it tricky to get an accurate representation of engagement. This is an issue for Public Affairs Officers (PAO) of the United States Army's Aviation & Missile Center (AVMC), my client for this project.
To solve this problem, I designed and developed a web-based application that automates the process of identifying and removing bot activity, significantly reducing the time and effort expended by the PAO. The resulting application is highly effective in capturing high-fidelity data, providing a more authentic picture of social media engagement
Timeline
Fall 2022
Role
Lead UX Designer
Client
US Army, Aviation & Missile Center
Class
Hacking for Defence
PROBLEM
Metrics skewed by bot activity hinder authentic engagement measurement and confuse decision-making.
Metrics skewed by bot activity hinder authentic engagement measurement and confuse decision-making. AVMC PAOs measure social media campaign results but face challenges due to inauthentic engagement from bots. AVMC seeks better ways to discern real engagement. Inconsistent metrics confuse senior leadership, hampering progress.
Example of bot activity
USER INTERVIEWS
The AVMC is forced to manually identify and remove bots on their social media platforms, which is a time-intensive process.
I conducted interviews with 13 individuals, including members of AVMC and representatives from other organizations facing similar bot-related issues with their social media accounts. I asked about their present encounters with bot activity, the challenges they encountered, their existing approaches for addressing these issues, and gained valuable insights into their daily responsibilities and utilized tools.
DATA ANALYSIS
I observed specific characteristics and patterns of bot activity in the data Excel sheets provided by the client.
📌 Lack of Personal Information: Bots often have incomplete or vague profile information, and their profile images may be generic or stock images.
📌 Mass Following and Unfollowing: Bots might rapidly follow and unfollow a large number of accounts to increase their own follower count, engaging in follow/unfollow strategies.
📌 Repetitive Content: Bots may repetitively post the same content, spamming the platform with identical or slightly modified messages.
COMPETITIVE ANALYSIS
There is currently no established protocol or software for effectively detecting and eliminating bots.
I conducted research on the current tools available on the market that attempt to address the problem of bot activity. Sprinklr, the tool my client currently uses, solely provides social media metrics and disregards bot activity. Other solutions, like the ones listed below, claim to filter out bot activity in engagement reports; however, they lack transparency in explaining how they achieve this or providing evidence of their effectiveness.
INSIGHTS
I created an affinity diagram to identify shared opinions, thoughts, and issues among PAOs, revealing key insights and common themes
💡 Users can recognize bot activity by certain indicative characteristics, fostering familiarity with such behavior.
💡 Facebook and Instagram exhibit higher bot activity than Twitter and LinkedIn, implying varying security measures or user verification processes.
💡 Users often take actions like removing, blocking, hiding, or reporting bot activity to uphold authentic online experiences.
IDEATION
I chose to create an Information Architecture to define the navigation, content hierarchy, and overall flow of the user interface
IDEATION
Progressing from ideation to wireframes, I facilitated meaningful discussions with the client, effectively translating ideas into tangible design concepts
Homepage
An analytics dashboard illustrating the bot activity on the social media sites.
Bot History
Table populated with bots detecting on social media sites. Ability to block, remove, or report activity.
Bot Settings & Preferences
Some users noted that they would like the platform to handle all bot activity whereas some users wanted to still be in control of how the bots are handled. Here is where they can select their preference.
Connect Social Media
Ability for users to connect multiple social media accounts.
TESTING
Speed-dated mid-fidelity prototypes with target users to rapidly gather feedback and iterate over concept
PROTOTYPING
Bringing the validated high-fidelity designs to life with HTML, CSS, and JS.
WIP! To practice my front-end development skills, I have decided to code up my designs after the course has ended. Follow my progress here: https://annarippert.github.io/bot-free/index.html or view my code on GitHub: https://github.com/annarippert/bot-free
REFLECTION
What I learned
👩🏼💻
Being able to collect and analyze data is key to creating better designs and user experiences. Users do not always know what they want, so it is important to be able to analyze all the information you can gather to help determine what the user really needs.
Keeping the client engaged throughout the process helps us make better decisions. I made sure to stay in touch with the client every week, sharing updates and asking for their feedback on the project's progress.
✍🏻