Reducing time to insight for investors through panel calls

Role

Design lead, Partial PM

Team

Design Intern, Engineering Manager, Director of Product

Timeline

6 Months, Q2/Q3 2023

Tegus

CONTEXT

Tegus is a research platform for investors that houses dense qualitative and quantitative content.

This project was to improve the experience for new qualitative content called panel calls. Panel calls are calls conducted by an investor with a panel of experts in a given domain. Our team was tasked with taking the transcript of these calls and adding them to our vast database of investor-led research calls.

What was wrong with panels calls?

Panel calls in a beta version accessible to 350+ users, was developed based on the existing 1-on-1 call format and process. However, this has revealed various challenges due to differences between panel calls and 1-on-1 calls, leading beta users to refrain from using panel calls.

The operations teams was working tirelessly to create an content flywheel.

Problem:

Beta users are avoiding using panel calls due to two primary issues—one related to product strategy and the other to design

  1. Investors are unsure how to effectively use panel calls in their research

  2. Panel call transcripts are hard to follow and not easily digestible (note below)

THE GOAL

How might we improve panel call engagement?

IDentifying opportunities

TURNING OPPORTUNITIES INTO SOLUTIONS

Utilizing an opportunity tree we identified potential solutions and created experiments we could use to test the solutions. We then kept track of those solutions and how we tested them.

TESTING

We conducted user feedback sessions  and a pop-up Hotjar survey on Tegus with the goal of understand what resonated with users.

12 total user interviews

350+ beta users surveyed

Here is our list of experiments from our tree above with the experiments we carried out along with some notes.

TESTING OUTCOME

Folks loved the buzz around upcoming panels, but creating context around multiple speakers and being able to consume long form content more easily resonated most with users.

documenting our findings

Here is our list of experiments from our tree above with the experiments we carried out along with some notes.

biggest pain points identified

Difficult to keep track of who’s speaking

forgetting expert details while reading

Lengthy and dense paragraphs make it challenging for users to consistently keep track of who is speaking and understanding an expert's background and perspective is crucial for interpreting a panel call transcript, but it's easy to forget.

using pain points to create design goals

  1. Ensure users remember the expert's background

  2. Make it easy to track who is speaking

ideation

Armed with our research and design goals, our amazing intern created these iterations.

implementation

We were ready to get our designs to development after our intern had left. We had two considerations when thinking about how to streamline our designs.

  1. Engineering Constraints: The way the backend was set up for documents added a lot time to implement the side panels and made responsive views more complicated

  1. Beta expansion: In order to get the best feedback before a GA release, we wanted to expand to an unbiased pool. This also gave us quick wins like timing of hover to be 1.5 milliseconds

OUTCOME

In the original beta, 15% users were reading panel calls transcripts MoM, by the end 40% of users were reading panel calls.

We released to GA and sent out a feedback survey to all users. A great way to sum up our work:

“This is so easy to read - you should do this to all your transcripts!” - Anji Ran, Hedge fund Analyst