information architecture | onboarding
My team met with Links.ee, a real-world company founded by Eric Peterson in Los Angeles, CA. Eric came to us with the problem that users on his site were dropping off and didn't seem to be using the full functionality of his site. He asked my team to create an onboarding experience to guide users through his features.
After user testing, we found that the real reason many users were abandoning the site was that they were confused of how to even use it, even after being instructed, because of a confusing site architecture and inconsistencies in design. We did thorough user testing and mapped where in the user flow users were having trouble so we could reconstruct the architecture of the website. After we built a stronger UX foundation, we then designed and implemented the desired onboarding, which now needed fewer instructions due to a redesigned flow.
Note: Links.ee used to call folders "maps". In the following documentation, the phrase "maps" refers to the folder the links are organized within. After user testing showing few people understood this, Eric changed it officially to "folders".
PROCESS - INFORMATION ARCHITECTURE
Our user research helped us understand how people were already interacting with saving links and organizing them. It showed us that Links.ee needed to assert why it was easier than the bookmark folder integrated in browsers and also competitors Pinterest and Dropbox. It also supported the idea that people want to share links, and their behavior of emailing links to friends or tagging them in social media validated that desire.
Eric provided us with personas of the people using his website to help us figure out what would be most important for onboarding. This also helped us eliminate some of the data from our survey of people that had fringe preferences. The main takeaway from the personas and research was that people were focused on organization, the ability to share, and to control their own settings. We focused on those needs first as they were important to many people and also were the foundation of the site. If they did not work correctly, none of the extra features that set Links.ee apart would be useful.
We did a task analysis with four users that fit our persona profiles. We gave them the task of creating a new folder and manually adding a link to it, filming their interaction on the screen. We documented this by taking screenshots of each page that was visited and marking with icons and notes where the problems arose for each user.
When we came together, we compared notes and tallied how many users ran into the same problems. Not all users took the same path, so even the roadblocks with just one user were looked at. From here, we were able to prioritize issues that all users had and those that would block the flow of completing the task. This is where we realized that the website wasn't ready for onboarding, yet. With only 3 weeks to work on this project, we really focused on the core functionality of the site - adding a folder and adding links. Once that foundation was set, we could then move on to onboarding. It didn't make sense to onboard users to a flow that was inherently confusing, especially when our user testing showed teacher people in the current framework still didn't help them accomplish the task.
We created a prototype for the new feature of being able to edit folders from the homepage. We chose to do an ellipses that when clicked would reveal options such as move, edit, and delete.
[images coming soon - please check out the prototype to see the final design!]
PROCESS - ONBOARDING
In the a survey that 63 people responded to, we were able to make strong decisions on how to move forward with onboarding. We found that it would be best to make sure the bookmarklet was not required to create a Links.ee account, but to reiterate the value statement and ability to download later. We also found that people would want to be able to view public or shared folders without making an account, where originally the site made you register an account first. People were okay with creating an account to interact. This showed people wanted to see the value in the site before they went forward, which we realized we could incorporate in the onboarding
We also learned what onboarding preferences our respondents had. The preferences leaned toward hints as you move throughout the platform. Based on competitor research and the need to showcase Links.ee's unique place in the market, we thought a short introduction would be useful where users could also download the bookmarklet.
We developed out the introduction to Links.ee in Balsamiq. During user testing, people were confused if the folders below "Mike" were something you could interact with, so we made it partially opaque in a later iteration. We also wanted the "Skip to Homepage" to be more prominent as people missed it in our first iteration.
We also integrated the user-preferred tool tips to teach people the core functions as they moved through the website. During user testing, users immediately interacted with the functions, validating its success as a way to onboard.
For next steps that Links.ee could take, we looked to what areas still needed iteration. There was a lot of confusion during user testing with inviting. Users were unsure when typing out names if they were searching within Links.ee and how to connect outside of it. We found that Pinterest had a good way of showing exactly where contacts were being pulled from that Links.ee could take inspiration from.
We also had feedback during testing that people were confused with all of the different privacy options. It seemed like there were too many and that overall the major components were whether or not a folder was publicly visible, and if not, who all was allowed to see it. Pinterest again designed for this same idea but only had toggle to make a board secret or not. If a board is secret, then only collaborators can see it. Links.ee could simplify their privacy settings as well to make their flow easier for users.
We delivered 18 iterations that were validated in user testing. In our user testing with the initial website, 0/4 users were able to go through the task of adding a new folder and link. After our iterations, 6/6 people successfully got through the task.
By utilizing the UX changes we made, the core functionalities of the website were addressed. This built a strong foundation for the other features like sharing, commenting, and following. Before, users were required to learn a flow that wasn't intuitive, and with all of the competitors in this space likely would have abandoned the site for good. Users now have a more intuitive experience and with onboarding understand that, unlike other sites, they can follow public folders and share their links.
Sketch 3 | Invision | Google Surveys| Marvel | Asana | Trello | Slack | Omnigraffle | Paper prototyping| White Boards | Post-Its
User surveys; Excel PivotTables to find correlations; competitive research; onboarding research; user flows; lo to hi-fidelity mockups; stakeholder presentation.