Linkedin Design Sprint

Challenge

Sometimes, when a user is happy in their current position at work, they choose not to spend their time on LinkedIn. I set out to design a creative solution to gain a dormant user’s attention who is returning to LinkedIn and solicitates an action from them to actively particpate on LinkedIn in anyway.

Timeline: 5 days

What does success look like: There is an increase in the amount of users engaging in any activity after returning to LinkedIn after 3+ months. There is an increase of users that convert from dormant to active users for longer than a 6 month time frame.

What is out of scope (a useful tip I learned from Slack): Due to the constraints of this challenge, I decided to focus specifically on the U.S. market. While I’ll keep in mind core principles for designing for internationalization, it won’t be my main focus.

Inside The Organic UX Design Process At Slack

Plan:

Thursday: Create Research Plan + Gather Participants

Friday: Conduct Research + Synthesis Data

Saturday: Rapid prototyping + Sketching + Wireframing

Sunday: Prototype with real people + Modify Wireframes + Visual Exploration

Monday: High Fidelity Comps + Write Medium Post

Research:

I wanted to first look at analagous experiences, so I brainstormed any situational event I could think of:

From here I pulled experiences that I thought would be the quickest to evaluate in the time frame that I had. I chose Airbnb, Amazon, and Zocdoc.

AirBnb Key Screens

Key Notes:

1. Primary activity user is logging in for is front and center, along with updated value prop (now includes experiences)
2. Call out for what is useful for you specifically
3. Call out for what is popular
4. Incentive to gather other users and to come back and use the site when you get your $95
5. Incentive to incorporate Airbnb into your worklife
6. Incentive to particpate in Airbnb outside of the UI (LA beta)

ZocDoc Key Screens

Key Notes:

1. The primary action the user is logging in for is front and center — combined with roatating scenarios + contextual images
2. Welcome back experience for your doctor, outlining steps (the user knows this won’t take too long) and allowing them to skip this part
3. Gives user the ability to continue booking appointments or rebook previous doctors
4. Highlights a wellness guide that calls user’s attention to gaps in their health care

Amazon Key Screens

Key Notes:

1. Search is not as prominant as ZocDoc or Airbnb, but Amazon remembers where you left off and balances the hierarchy between the three (search, shopping cart, recommendations).
2. After you complete your task (purchase), you have a confirmation of completing that task successfully with key info about it. You also see what is related to your order and past searches.

I then gathered particpants for a reasearch study by asking my network to connect me with people who they thought might be a good fit. The only requirements were that they had a LinkedIn, weren’t a power user, had displayed dormant behavior, and we have never met.

I opted for qualatative over quantatitive data for a few reasons. First, I wanted to explore, explain, and understand the “what” and the “why” of the dormant user’s behavior. Since this is a 5 day solo sprint, I also believed it to be imporant that the research plan be flexible — allowing it to evolve and change course as need be.

The group consisted of 4 participants (2 women, 2 men; 21, 20, 33, 26; nursing student, writer, ad operations, and marketer). We began with a group ideation session. I posed 3 questions to the group and each participant wrote down as many answers as they could think of within 2 minutes. Here are the results:

Question 1: What actions do you currently take on LinkedIn?

Question 2: What actions do you currently take outside of LinkedIn to advance your profession, build your network, or improve your skill set?

Question 3: What actions do you want to take to advance your profession, build your network, or improve your skill set in the future? (This does not need to be contrained by reality! Anything is possible!)

After we were done ideating, I asked the team to group these activities together based on common motivations and to label their groups. This helps identify what the underlying values are for professionals. I would extrapolate these values from the research study after the session completed.

Here is the final results of the groupings done by the group

After this, we moved on to drawing experience maps of the last time they were on LinkedIn. I asked them to be specific and aim for details. I did one as an example using the base activity “watching TV” to help illustrate the expectations of this exercise. Their responses helped me contexualize the flow I was trying to create. I wanted to understand four key aspects: when are people coming to LinkedIn, how they feel about the experience, why they leave, and what they do immediately afterwards.

Takeaways:

• People in my focus group all came in through e-mail prompts.
• People only began the flow feeling negatively if the person trying to connect was not someone they wanted to connect with. This means that users see LinkedIn as a neutral/positive tool to accomplish their task.
• People leave immediately after accomplishing their task. They go in with a goal to achieve.

Rapid Prototyping, Wireframes, and Usability: Go!

I began with sketching out some concepts and flows:

A few examples of the wireframes and flows I sketched out

I then turned these into quick and lo-fi rapid prototypes. These were printed out and used as a paper prototyping session for speed’s sake. This time, I asked for volunteers in a co-working space that a friend of mine works in. I aimed to at least match the number of people who participated in my focus group. People are always surprising with their kindness, and I was able to conduct 8 paper prototyping sessions. Each particpant was asked to enter LinkedIn through an e-mail prompt to connect with someone, ensure that action was completed, and then either (1) identify something that they’d find useful to their career and begin to take action or (2) connect with someone they admire and want to learn from, within three different potential flows. (These tasks alternated between participants.)

Here are the key screens of each flow:

The overall feedback from the participants was that:

The FIRST prototype didn’t offer enough positive feedback that what they set out to accomplish (connecting with someone) was achieved, but they appreciated seeing new people to connect to that they normally don’t see suggested to them (thought leaders). This peaked their interest as to what content these professionals were contributing to the online community.

“I didn’t know that you could follow people on LinkedIn — I thought you could only connect with people you knew.”

The SECOND prototype performed the weakest. People agreed that they felt disoriented coming into LinkedIn, expecting to see the flow that they’re used to (even if it has been some time). Being taken out of that flow to perform tasks that they didn’t feel they had enough time to complete in the moment left them feeling negatively.

“I just feel that I know why I’m coming into LinkedIn and what I want to do on here. I don’t really need help with that.”

The THIRD prototype was less intrusive but didn’t offer enough context of what exactly the user was getting. It seemed to be repeating the top navigation that they’re used to seeing on LinkedIn. They weren’t sure what they were getting that was “new.”

“This seems like what I’m used to seeing through notifications.”

* Disclaimer of the limitations of this research: This focus group served to give insight into some of the users of LinkedIn who have displayed dormant behavior. The group was not meant to be descriptive of the average user of LinkedIn. Because of the time frame and resources available to me, the particpant study consisted of people who’s lives look similar to some degree (i.e. they all live in SF, are familiar with tech, etc.). Because our friends tend to be like ourselves (and therefore the friends of our friends) — the study could not be completely random. By having that degree of speration, however, I was able to ensure some diversity and remove bias. This research proivded a useful tool in creating a dialogue around values, behaviors, and garnered enough insight to begin a larger discussion or project on how to engage users returning to LinkedIn after 3+ months. Ideally, the next step would be to create a research plan that reached a much larger and more diverse group of users.

Results are in!

Flow when accepting a connection from e-mail notification

Flow when responding to a message on LinkedIn through an e-mail notification

The salience principle tells us that a person’s attention is drawn to that which is most relevant at that moment. We can see an example of this above in Amazon’s flow — each view has suggestions based on their most recent searches or what they’ve previously added to their cart. It’s important to meet people where they are when they’re re-entering a site — and rather than offering a flow that they could drop out of (with a negative feeling), I wanted to put infront of them exactly what they are looking for.

I layed out four main boxes that reflect the values I saw in my focus group. They are limited to three suggestions at a time, that way a returning user doesn’t feel overwhelmed with tasks and rather can focus on what they want to achieve in LinkedIn. Since they are already feeling accomplished by completing a task, I opted to design a page that utilizes the motivation underlying their initial action.

There were three values that I observed in the ideation process and decided to focus on, making them actionable:

I want to advance my career

I want to grow my skillset

I want to learn with and from others

By putting the value proposition of each of LinkedIn’s offerings front and center, we’ll be making it easier for the user to identify what matches their intent in that moment and to take direct action. It was also important to make sure that the copy reflects that motivation. By focusing on simplifing what the welcome back experience could be, I was able to increase the likelihood of a user taking action and, even more so, of the user returning again.

Each of these four actions helps to customize their dashboard and e-mail notifications a bit better, which will help garner more activity and less dormant behavior. They also connect to LinkedIn’s revenue / business goals, by surfacing features of the tool that are sometimes hidden. By grouping content by motivation, we’ll be able to disperse multiple forms of content within the flow, displaying the diversity of LinkedIn’s offerings. This allows us to solve for the problem the product is trying to solve, rather than focusing too narrowly on one prescribed solution. It also increases the chances for the user to have an “aha!” moment by identifying something useful to them. We want to remind them of the core value of the product as quickly as possible.

Visually, I included LinkedIn’s pattern of providing social proof. This is important, particularly on a page where I wanted to utilize white space, because it’ll be easier for users to identify images that they already know. Plus, of course, it provides validation that the proposition is valuable or trustworthy.

Since I included a lot of information on the page, I decided not to have anything hidden below the fold. This was important because I wanted users to be able to quickly figure out what to do. We know that UIs that aren’t intuitive enough for users to figure out on their own tend to become frustrating and lead to users quitting earlier than desired. Since the content was busy, it was important to keep the UI clean, straightforward, and evenly distributed. This creates a symmetry and affords the user a sense of balance. However, there is still a hierarchy to the page allowing the user to move from higher level learning to career development.

Conclusion

What I’d do next: I’d love to understand, from a metrics perspective, what quantitative research has been done to better understand a broader set of behaviors users engage in when returning to LinkedIn after a period of time away. I’d also conduct a digital prototyping session with a larger group of participants. I’d then focus on honing the micro-interactions and figuring out what information is the most valuable within these categories.

For now, I am surfacing 3 categories because studies tend to show that people are more likely to choose 1 option out of 3 than out of other numbers. However, these situations are unique and I’d like further proof of what LinkedIn users find useful.