Usability Testing A New Memory Aid Feature
In 2019, I worked on a team to design a conceptual feature for the LinkedIn mobile app. We wanted to ease a key frustration:
“Sometimes I’ll just be scrolling through my feed and be like ‘where the hell did I meet this lady?’ I don’t remember her at all.”
Based on initial user research, we hypothesized that if they could add tags and personal notes to their contact records, users would have an easier time finding and remembering people in their network.
Prototypes and usability testing drove many of our design decisions.
DEMOS
add new contact (gif)
tag multiple contacts (gif)
filter tags to find contact (gif)
ACTIVITIES
For this project, I collaborated with two other UX designers. Together we conducted research, developed a design strategy, drafted sketches of solutions, and prepared prototypes for testing and demonstration.
My Responsibilities
Research Protocol
Business & Competitor Research
Research Report
User Scenarios
Future State App Map
Future State User Flows
Usability Protocol & Testing
Project Presentation
Tools
Paper & Pen
Axure
Sketch
Invision
Time
2 weeks
RESEARCH
LinkedIn claims to be the “world’s largest professional network,” with more than 630 million users in 200 countries.
Professionals are LinkedIn’s primary and most valued users. Their data is the product the company sells to its other customers.
Users average 17 mins/month on LinkedIn vs. 2+ hrs/day on social media in general.
LinkedIn wants users to engage the platform consistently.
LinkedIn has identified that as a member’s network grows, it can be difficult to remember and properly associate all of the connections.
Goals of research
Verify that users are struggling to remember their connections.
Understand what users want to be able to do, why, and what common memorable attributes they use to recall connections.
Understand how other platforms support connection recall.
What would make LinkedIn more valuable to users?
User interviews revealed that professionals who use LinkedIn:
Struggle to remember how they know some of their contacts.
Have difficulty finding contacts if they don’t recall the contact’s name or company.
Wish they could add relational information at the time they make a connection.
Want a way to understand the history of their interactions with their connections.
STRATEGY
Based on the research, we hypothesized that by providing “professionals” with additional contact management features, users would have a way to better remember their contacts, engage the platform more consistently, and — as a consequence — value LinkedIn more.
Our competitive analysis suggested a few possible solutions to pursue from adding a single simple note feature to turning LinkedIn into a full-blown contact management system. We knew from our research that users did not want LinkedIn to become more onerous or require lots of time to deal with, so our question was:
What would be the simplest way (from the user perspective) to add a memory clue to a network contact’s record?
As an initial step, we decided to offer two new linked features:
Customized tags, which could be categorized with some of the same major groupings LinkedIn uses to categorize Experience on a user’s profile — Work, School, Events, Projects, and other — to make integrating the new tags into existing search, sort, and filter tool seamless.
Personal notes, which would give the user further opportunity to go into more detail with attributes or other memory assistance information.
PROTOTYPE & USABILITY TESTING
usability test protocol (pdf)
We conducted two rounds of usability testing, first with a paper prototype and later with an interactive digital prototype. Via testing, we wanted to:
Verify LinkedIn users found the user flows we had developed intuitive.
Validate that our new features functioned the way users expected them to once they discovered them.
Gather heuristics feedback on the new features and their integration into LinkedIn’s app design.
Allow users behavior and desires guide certain design decisions, specifically:
What should the “tags” feature really be named – tags, labels, or something else entirely?
Did users want LinkedIn to auto-suggest tags or just create their own from scratch each time?
Did users want to view and sort their tags as a way of browsing their connections?
USABILITY TEST FINDINGS
User testers agreed with original interviewees that:
Organizing actions must be very easy and part of a current flow.
Users are too busy to organize contacts after they are a confirmed connection.
Once users hit a critical mass of connections, they want to organize their contacts to find and remember them more easily.
Testers did not agree about:
How much contact record information they wanted.
Where they wanted to access the contact record information either integrated into the highlights section on their connection’s profile page or on a separate page.
What “tags” should really be named – tags, labels, custom groups.
Given the small scale of our usability testing, we ended up making final design choices for these issues based on what we, as designers, thought might be best received. To actually determine the “right” solution to these kinds of usability questions, LinkedIn’s 630 million user-base could provide ample ground for A/B testing multiple options.
Testers consistently:
Liked accessing both tag and notes features during the “connection” flow and from the Profile Page of a contact.
Liked being able to see their own tags and shared attributes on the results/listing screens.
Liked that they could create custom tags that represented experiences, events, and other attributes that they or their contacts are not publishing on their profiles.
Wanted the My Network and Manage My Network pages to be merged into one page.
Did not see the value in LinkedIn suggesting “experience” or “education” tags that the system already highlights. However, users did see value in being reminded of tags they had created for other contacts, in case there was an attribute they wanted to apply to multiple contacts.
Did not want to batch-tag multiple connections simultaneously on the mobile app. They expected and wanted to search contacts one-by-one and add the tag, because on the phone doing anything but search requires too much scrolling.
ITERATING BASED ON USER FEEDBACK
Many of our iterations from paper through the multiple digital versions were smaller, subtle changes in copy or adjustments to layout. One key example of our iterations can be seen in the “tag tool” screens.
Our first tag tool draft was one screen that invited users to ‘tell their future self’ something via a tag but without a custom note feature.
It was clear we’d need more than one page to select the tag category, create the tag, add a custom note, and confirm for the user that their “goal” had been achieved, as well as more straightforward microcopy.
In the final version, we continued to refine the microcopy. Based on user feedback, we also removed the keyboard from second screen’s initial view. This simple step decluttered the experience for users, making them feel less overwhelmed. So while it added more screens, once the user committed to tapping into the text field, they were mentally ready for a keyboard to appear. Having the auto-fill suggestions appear next seemed to make the process even easier for them.
FINAL USER FLOWS
REFLECTIONS
This project was particularly enjoyable because it was such a collaborative effort. As researchers and designers, we each brought different ways of approaching the problem to the table, and it was in those differences that debate and inspiration arose. While our initial user research was limited, our usability testing provided key insights that informed both the functionality of our new features, as well the flow with which we moved users through the tagging and notating process. I was left though with one nagging thought.
Has the search function become so dominant that filters and tags are superfluous features?
Looking back on those usability tests, I was reminded of a moment when I asked each user, “How might you go about finding a person you know you have tagged, but whose name you have forgotten?” For each of them, their first impulse was not to use a filtering option to find everyone who shared the same tag, but rather to go straight to the search bar to put their tag name into the search field.
It made me wonder if the whole “tag” feature was adding an unnecessary and complicating layer. Search results could simply pull up a list of all the connections that have, for example, “Creative Mornings” somewhere in a custom notes field or anywhere else in their profile.
If users are being trained to use the search bar, not just in LinkedIn, but also in other commonly used apps and e-commerce sites, then perhaps all we need to give them is the ability to add a custom note that is editable and searchable. This solution might also be much easier to implement from the database side as well, as it only involves creating one new field per contact relationship rather than multiple.
Furthermore, as an MVP, LinkedIn could use a custom notes field to test whether or not their users actually have an appetite for using the platform a bit more like a contact management system as a precursor to adding features like tagging, labeling, or categorizing.
