It’s rare to be in a Zoom or Teams meeting these days that doesn’t have an AI note taker in the room. Sometimes, there are dueling note takers. Well here at Matrix Group, we have standardized on Fathom. For now. Here’s how we made the decision:
Note Taking is Hard
My Directors, Project Managers and I are in a LOT of meetings. And for every meeting, someone is tasked with taking notes. We take note taking very seriously at Matrix Group. I like to say that “she who takes notes, writes history.” And it’s true. If there’s ever a question about a discussion or a decision, we go back to the notes. Yes, today, we can go back to the recording, but that’s time-consuming and we don’t always keep the video or audio recording of a meeting.
Note taking is so important that we train our managers on how to take effective notes. We cover things like: how to format your notes, when to document the details and when it’s okay to summarize, how to capture all of the next steps.
The problem is, note taking is hard. You need to be a fast typist and you need to be accurate. After each meeting, you need to clean up your notes, send them out, and then parse the notes into tasks for team members. If a Project Manager is in back to back meetings, it could be a day or two before the notes go out. We decided there had to be a better way.
Putting AI Note Takers to the Test
Over a couple of months, anyone who was interested in trialing AI note takers took part in a test. We researched a bunch of note taking apps, assigned them to people, and told them to try them out. We asked them to:
- Check the accuracy of the time stamps
- Check the accuracy of the transcript
- Check the accuracy of the notes
- Explore the types of note formats available, e.g., summary, detailed, etc.
- Check for the ease of turning on and off
We tested a bunch of apps, including Fireflies, Otter, Tactiq, ChatGPT, and Fathom.
We Chose Fathom
After a lot of testing, we chose Fathom for the following reasons:
- We carefully reviewed the Terms of Use and were satisfied that client confidential data that might get discussed during a meeting would not go toward training a public LLM (large language model).
- The free version gave us all the functionality we needed.
- The notes are terrific, especially the summaries of meetings that wind around and come back to topics.
- The transcript is accurate.
- Fathom largely ignores the irrelevant chitchat at the start of each meeting.
- We liked the flexibility that Fathom offered. We could have Fathom always join, join only when requested, start and stop recordings, and connect to our calendar.
Yes, we chose Fathom BUT we are constantly testing out different tools. There are new AI note taking tools that pop up every day and while we’re not going to jump at the slightest chance, we are monitoring the landscape. If a better tool comes along, we will make the switch.
Our experience with Fathom by and large has been terrific. Project Managers report being able to get the notes out to clients much faster. Aside from one weird day when a set of notes, instead of talking about a website redesign, discussed the Three Little Pigs. Who knows? Maybe Fathom was reading some fairy tales over lunch one day. The next meeting was fine.
In the next blog post, I’ll talk about HOW we are being intentional with our use of Fathom, training Fathom AND training ourselves to work better with our AI note taking tool.
How about you? Is your organization using an AI note taking tool? What has been your experience? Why did you pick that tool?
