Proprietary professional network redesign
Alchemist Accelerator is a startup accelerator based in San Francisco. Access to an extensive network of faculty, mentors and investors is one of the valuable perks provided to the founders of startups accepted to the Alchemist program.
User Research, Concept Ideation, User Testing, Wireframing, Prototyping
1 Product Manager
4 Software Engineers
1 Designer (Me)
Analyzing founders' support requests I noticed that many of them were related to the inability to get satisfactory search results. It was strong signaling that something was wrong with the existing filters on the platform. They were either not noticeable enough, or were not allowing founders to get desirable search results. I needed to figure what exactly was the problem and how it could be improved.
How could we improve on users' current search experience so they can find a mentor, coach, investor or another founder to reach out to and so that this person is a perfect fit for them?
Old Version, List of People
Old Version, Filtering
In order to identify usability issues and to discover users' needs that haven't been addressed yet, I have conducted 5 contextual inquiry sessions and 15 usability testing sessions.
contextual inquiry sessions
usability testing sessions
Contextual Inquiry Sessions Takeaways
During contextual inquiry sessions, I asked people to describe and demonstrate their typical search process. It helped me to identify what they prioritize when browsing through the people list.
I would say I typically use search by typing in a person's name if I know whom I'm looking for specifically. Otherwise, I search by typing in a term like "angel" and then just start browsing and seeing who's a good fit, just scroll down and look at each one.
I don't use many of the filters because I'm afraid that I'll miss someone good if I put on a filter. Part of that is because I don't know when someone new was added or updated. If I could see a date for each profile, that'd be helpful I think.
From contextual inquiry sessions, I discovered 3 major use cases:
1. Specific Person
User looking for a specific person by typing a person's name into the search bar.
"Do they have Walter Jameson from Pine VC on the platform?"
2. Group of people
User browsing through a particular group of people.
"I need to find a fundraising mentor with expertise in AI."
User applying a set of filters to narrow down the results.
"Are there any pre-seed angel investors in SF focused on the AghTech? "
3. Set of filters
Usability Testing Takeaways
I have conducted 15 remote usability testing sessions. I observed users performing a list of tasks to discover usability issues we were not aware of. After completing the sessions I wrote down all of the observed usability issues and began to cluster them based on the topic. As a member of a remote team, I've created a digital version of my findings to share it with the team.
Clustering usability issues based on the topic
As a result, there were 4 major groups of issues that emerged out of this clustering process:
Frequency of Use/Familiarity - some of the icons were not familiar to the participants. Also, I've noticed that they spent most of their time in the same menu section and didn't use the left sidebar frequently.
Filters - there were a lot of issues with filtering such as inability to make multiple selections or getting 0 results with too restrictive filters.
Search Results - some information important for the participant, such as profile rating, was missing in this representation. Also, there was not enough flexibility in the sorting process as well as adding to Favorites.
Updated/Outdated Profiles - most of the participants mentioned that they struggle with finding newly added profiles and that it's hard to understand whether a profile is up to date or not.
Competitive Heuristic Evaluation
From user research, I knew that a majority of our users were LinkedIn, Angelist and Crunchbase "power users". It led me to explore these products' filters and search results layouts to discover the relative strengths and weaknesses of their designs.
# of search results for each option in the filter helps to identify the most popular options
All filters are expanded, so it's easy to find the one you need
Some filters do not fit above the fold
No feedback if the option is not found in the list ("ecom" example on the screenshot)
Some filters have limited # of options without any obvious reason (i.e. Location).
Ability to Save search settings and to create alerts based on them
Prominent labels of the selected filters
The same options can be selected in different filters (i.e. Big Data on the screenshot). It's not obvious what filters are activated though.
The only way to figure out which columns are available for sorting is to hover the name of the column.
Big variety of filters and very flexible way of applying them
Tabs names have icons to improve scanning
Filters may be intimidating for less tech-savvy users.
Applying multiple filters significantly decreases screen space with search results.
Identifying these strengths and weaknesses helped me to crystalize my vision of the improvements I wanted to implement and allowed me to avoid some possible mistakes.
Empowered by user insights and competitive heuristic evaluation I started to explore potential design directions on paper and later in Sketch.
Sketching different layouts
Design exploration in Sketch
User Testing of the Two Filtering Options
After exploring multiple different layouts I ended up with two final options.
All filters with the most popular options for each filter are visible at one glance.
Batch filtering increases chances to get to zero results and makes it unclear which filter among the applied is the most restrictive one.
Expanded filters cover the list of results.
Some filters do not fit on the screen. User has to scroll down to see them.
Results are updated after each applied filter which significantly decreases chances of getting zero results and makes it obvious which filters are the most restrictive ones.
The list of search results is always visible.
Expanded filters use a decent part of the screen space which can be a frustrating experience on the smaller screens.
To identify the best between these two options I've tested them both with 2 different groups of 5 users each. I asked them to find Angel Investors, investing in Pre-Seed and Seed rounds and interested in the participant's industry.
Testing Option 1
👩🏻💻 Number of participants: 5
⏱ Avg. time on task: 2 min
🌟 Success rate: 100%
Testing Option 2
🧑🏽💻 Number of participants: 5
⏱ Avg. time on task: 1.5 min
🌟 Success rate: 100%
Based on the conducted usability testing sessions Option 2 performed better:
Time on task KPI was better than in Option 1.
It also provided more flexibility in avoiding zero results because search results were updated after each applied filter (in comparison with a batch filtering in Option 1).
Collapsed left side bar
Narrow left sidebar expands on hover.
The search bar is now smaller and is combined with the most popular filter.
A combination of checkboxes and Add Selected to List button allows adding multiple people to the list.
"New" and "Updated" tags help to illustrate that the network is constantly growing.
Filters are grouped into three lines with icons. The same icons are used in the applied filter labels.
Based on user feedback I added new filters related to the investment characteristics. I created a separate group for these filters.
Search Results screen after redesign
Please use arrows to see the Before screen
Filtering after redesign
Please use arrows to see the Before screen
During the user interviews, I've discovered that the most popular use case on mobile was #1: Looking for a specific person by typing a person's name into the search bar. Users didn't perform complex searches on mobile, they were using it to quickly check whether the person is in the database or not. Based on this insight I decided to go with an abridged version for mobile.
The most popular use case on mobile was searching for a specific person. That's why I decided to keep the search bar as an essential interface element and to remove Advanced filters and Add to List functionality on mobile.
This decision helped us to save a decent amount of time on designing and developing these features and users didn't miss this functionality on mobile.
Iterative user research and prototyping led to a final solution that significantly increased users' satisfaction with the product. Now users can perform search faster, it is more flexible and allows them to find a better fit.
We continue to receive positive feedback from our users.
"I'd say the current version of the product is a huge improvement from when we first joined Alchemist!"