Linkedin Elicits - Inciting authentic connections among young professionals
Design process
Design thinking
Timeline
72 hours in 2022
Team
Ethan Pemberton - UXR + UI designer + Prototyper
Alec Faeste - UXR
Zoe Janssen - UXR
Nick Gunning - UXR + UI designer
Project overview
This project was undertaken as part of the 'Canva x Prodigi Productathon'. We were given 72 hours to select a social media app and design an experience within the in-app editor that could potentially boost interactions and engagement on the platform.
The problem
LinkedIn is notoriously known as an outdated and superficial app amongst younger demographics. Thus we asked ourselves 'How might we create an in app editor feature that helps young professionals connect with industry professionals in a meaningful and authentic way, so that they may make advances in their career'.
The solution
The solution we arrived at was 'LinkedIn Elicits'. Elicits engage the younger demographic to connect with others by creating daily elicits that encourage users to participate in authentic conversations and allow for more mutual points of connection amongst professionals. This feature helps lower the barrier to entry for younger users to connect with more senior professionals.
Design process
With only 72 hours to research, ideate, and design a feature, we knew we had to ensure our choices of tools and methods were intentional and impactful. We followed a design thinking process to keep the potential user at the centre of our decisions and ensure we would create a feature that would genuinely help and create a better experience for out target demographic.
Empathise
Questionnaire
Netnography
Competitor analysis
Define
Personas
Affinity diagram
Ideate
Crazy 8s
How might we statements
Prototype
User flow
Wireframing
Figma prototype
Test
Usability testing
Think-aloud protocol
At a glance: Research
Efficiency and impact were the main drivers of our research methods due to the time constraints we had. To guide our research and ensure we would retrieve the right information, we define four key research questions:
What are the patterns of engagement and primary uses of LinkedIn among users aged 18-25?
What objectives or professional goals do LinkedIn users within the 18-25 age group primarily seek to achieve on the platform?
What common functionalities or features on competing social platforms have been identified as significant drivers of user engagement?
What are the primary pain points and areas of dissatisfaction for users when interacting on LinkedIn?
Great design doesn't live inside designers. It lives inside your users' heads. You get inside your users' heads by doing good UX research
- David Travis
Research findings
Understanding motivations
We distributed a questionnaire to various online communities on Reddit, as well as among our own personal connections. The main goal of this survey was to get a clearer picture of why people use LinkedIn. In addition, we asked questions to understand the typical activities users engage in on LinkedIn, and to identify any frustrations they might be experiencing on the platform.
85% of users responded stating their experience with LinkedIn was very poor. Furthermore, 100% of participants stated they never spent more than 10 minutes on the platform in a single session. This does not say the whole story, as those 10 minutes may be 10 impactful minutes. But from a business standpoint, LinkedIn must work at keeping users engaged for longer periods of time.
Here were our high level findings from the thematic analysis of questionnaire responses:
Identifying common patterns of engagement across social media tools
The utilisation of the competitor analysis allowed us to explore various key features of competing platforms such as Snapchat, Instagram, TikTok and Twitter. In doing so we aimed to identify common patterns within the included features on each platform and see if this could possibly be translated to LinkedIn in some form to boost engagement.
We found that platforms that maintained a significant hold on user engagement had:
Understanding behaviours and attitudes towards LinkedIn
We used netnography to immerse ourselves in diverse online communities, with the aim of understanding the common challenges young users face on LinkedIn. By reviewing discussions and comments on platforms like Reddit and Apple App Store reviews, we gathered a variety of opinions about LinkedIn, both positive and negative. We then conducted a thematic analysis of these quotes to identify recurring themes and key insights. These findings included:
Defining the problem
Aiming to answer the question 'Who are we designing for?' We synthesised our research in the form of user personas. By understanding the expectations, concerns, and motivations of these users, it’s possible to design a feature that will satisfy users’ needs and therefore be successful.
Overall findings
To get us all on the same page and to gain a deeper understanding of the research we conducted, we used affinity diagramming. This method got us all involved and discussing the research and pulling out user needs and findings together so when it came time to defining the problem we could all be confident. Overall we found:
User personas
We used personas to help us stay anchored on the users and to avoid letting our desires trump user needs. For each user we identified primary frustrations as well as wants and needs. This was particularly useful when defining the problem as we were able to hold up the problem statement in light of these personas and see if it genuinely aligned with their wants and needs. Here is a brief overview of the two:
Problem definition
With all this knowledge and understanding, we developed a targeted and defined problem so that we could have clarity of direction in our ideation solutions. We landed on:
Young LinkedIn users have no meaningful and authentic way of connecting with other industry professionals and employers. Because of this, they resort to superficial cookie-cutter posting approaches to gain traction which in turn adds to the noise.
Ideating solutions
Creativity is allowing yourself to make mistakes. Design is knowing which ones to keep.
Scott Adams
Following this new design problem we employed the methods of 'How might we' (HMW) statements as well as Crazy 8s to rapidly ideate potential solutions.
How Might We?
Taking key themes from our research, we set them up as individual HMW statements. Through the lens of the newly defined problem we then set a time limit and rapidly pushed out ideas on sticky notes. This method allowed us to ideate solutions for multiple facets of the problem and whilst also allowing us to make mistakes or put out silly ideas that in turn could lead to a greater final idea.
Crazy 8s
Similar to HMW statements, Crazy 8s allowed for rapid ideation under a forced time limit. This method was different to HMW in that it was not limited to key themes but instead allowed for a broader scope of ideas which could solve the problem statement in adjacent areas. This would allow for a fuller ideation process which would create a more robust final solution.
Assessing these potential solutions
From these ideations we ran through a process of voting, with each team member being given a number of dots to cast their vote on a design. We also discussed how certain ideas could be merged together to create a more robust and effective solution. We chose to make the decision this way due to the time pressure. It was moments like this when the real value of doing the affinity diagram together was revealed, as we all understood the research to make an impactful decision.
Design exploration and testing
We decided to move forward with an idea of daily prompts being sent to LinkedIn users to respond to. This solution would address key findings we had in our research such as creating a natural connection point for users as well as pulling users away from superficial posts.
Designing a user flow
In order to truly understand what screens were needed and understand more of what actions and decisions users would take with this feature, we developed a user flow.
Low fidelity
With this understanding of the user flow we then proceeded to develop low fidelity wireframes for testing. These wireframes were done to allow us to quickly reach a minimum viable product that we could usability test with and therefore quickly iterate upon.
Usability testing
For usability testing we provided 5 participants 2-3 tasks to complete. We asked that they would think-aloud throughout the task and speak on anything they found challenging or strange. We then recorded their responses and highlighted any standout comments.
Finding 4
Users would often have a difficult time remembering what the prompt was when they went to respond
Finding 5
Users were still left a bit standed with how to connect with these professionals from these elicits.
Linkedin Elicits
Iterating based on feedback
Taking on this feedback from the usability testing we addressed and iterated. We specifically addressed the difficult time users had figuring out how they can take the connection further through these elicits as this was a fundamental user need from our earlier research.
Key takeaways
Everything must have a justified use
Given the time pressure, I had to rid myself of the thought of sticking to certain research and design methods. Usually in an assignment we are tasked with mandatorily incorporating a certain amount of specific methods. Thus, I had to break out of this and really hone in on what methods were actually needed and would make a significant impact. Going forward and into the real world, I will always keep in mind: ‘Is what I’m doing actually working towards understanding the user or improving the experience of the product?’ If not, throw it away!