digital ethnography
Carnegie Mellon University | 2022
This project was a 6 week-long project for a Digital Ethnography class.
The purpose of this project was to be able to research a digital community and find recommendations for a company using emic ways.
We had to find a digital space to research using ethnographic methods, find insights and create visuals to communicate our findings.
Team: Genevieve Johnson, Anne Milan, Tanvi Bihani

areas
methods used
-
Ethnography
-
UX Research
-
Design Thinking
-
Ethnographic Research
-
Observations
-
Interviews
-
Secondary Research
journals, articles
-
Culture Mapping
research context
We were curious to learn about the discrepancy between a) a person’s identity, b) the persona they project on LinkedIn, and c) the persona they WANT to project on LinkedIn. To define each of these terms:
-
A person’s identity — All of a person’s many personas, which represent their hobbies, interests, personal growth, ties to loved ones, and professional goals
-
Persona they project on LinkedIn — The tone and content of a person’s posts and interactions on LinkedIn
-
Persona they want to project on LinkedIn — The desired tone and content of a person’s posts and interactions on LinkedIn, usually directly related to professional (money-making) goals
Do users feel restricted by LinkedIn’s implicit social rules? What are their goals and main reasons for interacting on LinkedIn?
overview of domain
LinkedIn is the world’s largest professional network. LinkedIn’s app and mobile website offer entry points to this social network. However, offline, nearly any student we might approach at Carnegie Mellon probably has a LinkedIn account. LinkedIn caters to a highly educated, young professional user base. According to the Pew Research Center’s “2021 Social Media Use” study, over 50% of American adults with a bachelor’s or graduate degree are LinkedIn users, while only 9% of people who didn’t finish high school have accounts. LinkedIn’s users are working professionals — 44% earn over $75,000 per year, while the US median salary is between $47,000-$57,000 for women and men, respectively. Over 60% of LinkedIn’s user base are Millennials (born 1981–1996), 13% of whom work in decision-making positions.
the client
We're targeting LinkedIn's "Member & Customer Experiences" team since our findings will relate to the quality and quantity of their platform’s user engagement.
target audience

research plan
PHASE 1
Methodology: Analysis of Linkedin's infrastructure.
Goal: To understand the platform and its capabilities, the technology tools it provides, and the number of restrictions on what posts and interactions can happen
Questions:
-
What are all the mechanisms for interaction on LinkedIn’s platform?
-
How does LinkedIn augment human relationships?
-
How does LinkedIn support information sharing?
-
How does this platform’s architecture impact its culture?
-
What structure or restrictions does LinkedIn provide to its members?
PHASE 2
Methodology: Tracking the posts and interactions of 6 specific users
Goal: To understand the tools users on LinkedIn use and the interactions they have with different levels of connections on the platform.
Questions
-
Which reactions do users commonly use and which ones are more rare?
-
What are the kinds of posts that garner reactions and comments?
-
Is there a correlation between interaction type and the degree of connections?
-
Is there a correlation between interaction type and post type?
-
Can we tell the user’s interests through their LinkedIn profile?
PHASE 3
Methodology: Semi-structured interviews with these people, plus a few other LinkedIn users
Goal: To understand how people describe their personal identities and their goals vs how they behave.
Questions
-
Why did you make a LinkedIn account?
-
What parts of your profile did you choose to fill out? (bio, skills, experience, education, open to work, picture, etc)
-
What are all the different ways you use LinkedIn today?
-
How often do you open LinkedIn?
-
What subject areas are you passionate about?
-
Do you seek out content on LinkedIn about these subjects? How?
-
Are there subjects you’re interested in that you never see on your LinkedIn? How do you feel about this?
-
What makes you interact with a post?
-
Does anything hold you back from liking/commenting on a post?
-
Are there rules you feel you need to follow as a member of this community?
-
How long does your conversation or interaction last when you connect with a person on LinkedIn that you’ve never met in real life?
Usability Mechanics
From a usability perspective, LinkedIn offers a complex setting for members to connect with each other and search for jobs. Although not an exhaustive list, mechanics include:
-
Connections
-
Posting
-
Reactions
-
Search
-
Messaging
-
Following
-
Number of Connections
-
Groups
findings


Of the 6-person interactions sample we studied, 60% were original posts, 55% were ‘likes’, and 27% were ‘reposts’ with original commentary. Nearly half (45%) of ‘likes’ were in response to a post by a 1st-degree connection, and 20% were responding to a post by an organization they follow. These stats were exaggerated among comments, 62% of comments responded to a 1st-degree connection, and 25% responded to an organization’s post. When it came to the content of posts, 19% of ‘likes’ were for articles written by strangers and 16% were for event announcements. 28% of original posts were sharing a job opening, while none were announcing a new job acceptance, and 50% were sharing an original, short insight (though our data may have been skewed by the people we tracked; one participant shared more of her thoughts on LinkedIn than the average user).
the stats


what the stats mean
‘Likes’ are low effort; the majority of interactions on LinkedIn aren’t mentally taxing, despite the platform trying to encourage intelligent thought. Even though members are aware that every interaction helps reinforce their feed content by educating LinkedIn’s AI, they’re willing to sacrifice their curated feeds to support first-degree connections. People feel more motivated to respond with a high-effort reaction (commenting, reposting, anything other than ‘liking’) for first-degree connections. People usually use LinkedIn for a dual purpose: personal edification and job search.
insights
1. Professional ≠ phony
Members prefer that LinkedIn provides a platform for predominantly professional interactions. While LinkedIn’s infrastructure allows for any kind of post, its culture has developed a shared understanding that personal posts are taboo. However, members generally like this restriction and are mainly okay with this because they like that LinkedIn provides a platform for one specific persona (professional expression).
“Even though I see people post about personal topics, I know it’s a professionally-oriented community, so I don’t post my creative poems or writing. I like that there are rules; it’s a useful tool, not a social media. It preserves the point of the tool.”
2. Perception
A. Awareness of the pedestal: The fact that every interaction is publicly shared greatly impacts user behavior; people are more thoughtful about which posts they engage with because they want to reinforce the persona they’re trying to project.
B. Rule of perception: Posts on LinkedIn are meant for others. Interactions and profile creation are done to create a good perception of them for others.
“There’s one creator whom I want to comment on, but I know that everyone will see what I write… I really respect {her}, so I want to be careful what I write.”
3. Manipulating the algorithm
People choose whom to follow, what to engage with, and whom to unfollow based on the content they wish to see. They’re extremely aware that they’re engaging with AI, and they’re trying to make it easy for the software to push content actually related to their interests. Members like and share posts or follow people based on how they want to curate their feed. Some users wish to like content that they want to see more of, while others do not want to reinforce content by interacting with posts they like.
“I would devise my recommendation engine to give me this content. I try to follow people who are closely related to these interests. A few days later, I see a lot more from these recommendations. I’d like to say I have a say in what it recommends. I’ll like 5 tech accounts or PM accounts, just so the engine gives me more.”
4. Conformity to professionalism
People made a LinkedIn profile because they thought they had to. They were advised by educators, friends, or family to join the network.
“I joined the platform because my peers and family members had it. It felt like I had to create an account if I wanted professional advancement in my career. “
5. Journey matters
LinkedIn users wish to see more about the process of the journey, rather than the end goal. There is a level of authenticity that is lost on LinkedIn when it comes to sharing about achievements, as those are the posts that are usually liked by immediate connections.
“I wish I could see more of the process of building and creating rather than just the end of the journey.”
6. There’s no time like the present
There’s an element of timeliness involved. People only post about an achievement the moment it happened; they don’t share old news.
“I use LinkedIn like a live resume that I update often and that is useful.”
7. Friendship comes first
People use stronger reactions (love, celebrate) or even stronger reactions (comment, repost) when the original post was by a first-degree connection. They were self-aware of this behavior; it matched our quantitative analysis and their self-reflection.
“Insights, things that are new developments that are relevant for people in my network. I don’t comment as much as I’d like. I comment on people I know, say congratulations and stuff.”

recommendations
-
Personal Space: Tag personal content when posting, so people can filter feeds for professional or personal content
-
Matchmaking: Recommend these profiles of interested candidates for jobs
-
Selective Privacy: More privacy around the content they like or interact with
benefits to linkedin
-
More active users: With LinkedIn recommending profiles, users that currently do not put in effort into curating their profiles will have more of a reason to do so. This will also increase Linkedin’s chances of more members paying for a premium subscription.
-
Enhanced perception: LinkedIn can be seen as more user-friendly and customizable if it can respond to what the job seeing users want and care about.
-
Increased user base: By allowing customization, Linkedin can retain customers that are on the platform for different reasons. Now, people that like seeing personal posts will not be opposed to joining Linkedin because it is strictly a professional space and vice versa.