UserVoice founder Richard White's newest product was born out of Zoom fatigue and a realization that the best insights from his calls didn't stick. Here's how he turned his Zoom stress into Fathom.
That’s how many Zoom meetings with customers and prospects I did in the first 6 weeks of 2020.
20 to 30 minutes a piece, typically in blocks of 4, with a 10 minute gap in between each.
It. Was. Exhausting. But not for the reasons you might expect.
You see, I love talking to customers and prospects. I enjoy the challenge of coming up with the right sequence of questions on the fly that unlock insights and uncover hidden pain points.
What was exhausting was trying to quickly type up my notes while also playing the role of Sherlock Holmes.
On top of that my notes were unintelligible to anyone but myself, and if I didn’t flesh those notes out into something more “real” in the minutes proceeding a call they would quickly become unintelligible to me too. But I put in the work because I knew these conversations were important and, as notes go, mine were pretty solid.
So you can imagine how disappointed I was when I’d frequently refer back to those notes a mere couple of weeks later and ask:
“Wait, which call was this?”
Show Don't Tell
All of these calls were part of a research project at UserVoice
, the company I previously founded.
As much as I hated taking all these notes, they were essential to communicating to the team the things we were learning about the product we were developing.
Across those 300 calls were a lot of powerful, candid reactions that got me fired up that I was excited to share with the team. Imagine my disappointment then when the collective reaction to my research was “meh.”
Those amazing moments I had with our customers simply didn’t translate to the team in bullet point form. Frustrated, I decided to try something different.
For our next debrief, I wouldn’t attempt to relay what I heard. Instead, I downloaded all the most recent calls and painstakingly edited them down into a 3-minute highlight reel.
“We’ve had a number of breakthroughs from our latest research calls but rather than recite the takeaways, I’d like you to show you what folks are saying.”
The response couldn’t have been more different than before. There was a buzz after that amongst the team that was palpable. My excitement was now their excitement.
That got me thinking that maybe my note-taking process wasn’t as effective or as essential as I thought.
"Ok, But What Did They Say Exactly?"
At this point, I was starting to get pretty curious. Between the note-taking pain points that I felt personally, and what I learned from my experiment with the team, there seemed to be a real opportunity here.
However, this research project was also a pretty unique situation. I don’t typically do so many back-to-back customer calls a day, but it got me thinking about folks that did — sales and success teams. I had actually run our sales team for a brief stint a few years prior and like a flash of lightning, I quickly remembered the most challenging part of that role:
I could never get sales reps to write good notes.
Some reps wrote notes that looked like mine if I never edited them after a call — a random list of words. Other reps seemingly transcribed their entire calls making me worry how they had time to actually have a conversation.
No matter what end of the spectrum the sales rep was on, I always had two problems. The first was that I found it difficult or impossible to get them to adjust their note-taking style, in any direction. It was clearly a deeply personal process that had been ingrained in them over a lifetime. The second problem was that regardless of how good or bad the notes were, I found myself asking the same question over and over again as we discussed their deals.
THEM: “They seemed to hesitate when it came to pricing”
ME: “Ok, but what did they say EXACTLY”
THEM: “They had a technical question about _________”
ME: “Ok, but what did they say EXACTLY”
THEM: “They were confident they’d sign this month”
ME: “Ok, but what did they say EXACTLY”
I realized I had the same problem that my team had when they heard my research notes. The important stuff was the nuance not captured in the notes: the exact words that were used and the tone they were delivered with.
And I realized those sales reps faced the same problem that I had on my call. No matter how good or bad their notes were, it was hard to remember all that nuance when you’re on back-to-back calls all day long. Everything just kind of blurs together.
A Flawed Prototype
It was at this point that the idea for what would become Fathom
started to form.
We started with a prototype where you could import a Zoom recording and then pull out specific segments of the call as standalone video clips. We even transcribed the entire call to make it easier to search through and find the important moments.
It was an improvement over me creating highlight reels by hand but there were still a couple of key problems. The first problem was that I still needed to take notes, otherwise, I wouldn’t remember what I wanted to go back and pull out of the recording after the call. The second problem was that it often took 30 minutes or more for Zoom to make available a recording after a call had ended.
This was problematic because as excited as I was to go in and pull out the highlights right after the call ended, I was much less excited to do this work 30 minutes later once I’d moved on to other tasks. Although the output might have been better, overall this was worse than my original note-taking process because it felt like more work not less.
Sidenote: We’ve subsequently learned from product usage that only about 15% of a call is noteworthy so it makes sense that we wouldn’t want to rewatch entire calls or even scan through them to find the small portion that’s important.
It was clear that any solution would need to provide instant access to the recording after the call ended to match the immediacy of my note-taking process.
I remember the reaction the team had when I told them that it needed to be ready in seconds not minutes. I understood this was going to be no small feat. It would require us to build out our own real-time recording infrastructure, get clever with transcribing the call in parallel to that recording, and a multitude of other optimizations.
It was also clear we would need to minimize post-call “work” like finding and pulling out the noteworthy sections of the call. To that end, we added a notepad where I could take notes and have them tied directly to that part of the call recording, but this still required me to type while talking (which causes my mouth to stop moving). So we replaced the notepad with a series of buttons, like INSIGHT or POSITIVE REACTION, that I could simply click whenever I heard something noteworthy.
The final piece of the puzzle was the realization that we could reasonably break up these calls into segments of back and forths between host and guest.
So when I clicked INSIGHT instead of just marking that point in the recording, Fathom could figure out the start and end of the noteworthy segment by looking back to when the other person started talking and then listening for when I started talking again.
It wasn’t 100% accurate but it was good enough (especially once we tuned it to ignore short phrases like “uh huh”), and it was way less work than going back into the call afterwards to pick out the perfect start and end points myself.
Draw The Rest Of The Owl
There’s nothing quite as satisfying as solving your own problem. And within a few calls with this new solution it was clear we were doing just that.
But it wasn’t until the app first crashed on me and I had to resort to my old note-taking process that I realized how hooked I was.
Zoom calls without our app now felt unnecessarily stressful.
It was this alpha version of Fathom
that we applied to and were accepted to Y-Combinator a little over a year ago.
Since then we’ve expanded the product to automate away more post-call work (ex: automatically sending highlights to Slack or auto-generating call summaries and putting them into your CRM) as well as make running a Zoom call less stressful (ex: automated alerts about time remaining or if you monologue for too long).
Draw the owl, a value made popular by Twilio, means there's no guide to problem solving. You have to doing the hard and creative work to finish drawing the owl.
We’ve also spent countless hours building and rebuilding the backed infrastructure to be highly available and reliable (did I mention that real-time is hard).
Like any startup, there have been many ups and downs but it’s been easier to have the conviction to see them through because I know how this product has changed how I work and how my team shares what they’re hearing from customers.
Every day there are millions of customer conversations, now mostly happening over Zoom, in which there are a multitude of brilliant insights, ideas, and powerfully candid reactions. Too many of these moments are lost in the translation from human conversation to bullet points. Too many CRMs are filled with notes that miss out on the nuance that comes from hearing something first-hand.
And too many managers are asking “Ok, but what did they say EXACTLY” like I once did.
I’m confident we’re on our way to changing that. I’m excited for a world where people can focus on their conversations rather than their notes. A world where everyone in a company can see first-hand what customers are saying instead of playing a giant game of telephone.
I know we’re already making that a reality for ourselves and a small group of early adopters. I hope we can make that a reality for a bunch of Product Hunters too.
Thank you for making it this far! If you’d like to try out Fathom for FREE check out our Product Hunt page. If you have any questions or thoughts on the product or the product-making process feel free to message me on LinkedIn.