Doug Camplejohn
(00:01)
Hello everyone, this is Doug Camplejohn, and on this week’s episode of Revenue Renegades, I’m excited to welcome Richard White, the CEO and founder of Fathom Video, which is on fire. Welcome to the show, Richard.
Richard White
(00:14)
Thanks for having me.
Doug Camplejohn
(00:16)
As a fellow founder, I always like to hear founding stories first. Tell me a little bit about where the spark for Fathom came from and how you got started.
Richard White
(00:26)
Sure. I ran a company before Fathom called UserVoice. Near the end of my tenure there, we were a small team doing a lot of user research for a new product. This was early 2020, right before COVID. I was doing 10 to 12 Zoom meetings a day-maybe more-each about 15 minutes, rapidly trying to type out notes. As soon as each session ended, I’d try to clean up those notes so they made sense, but they were all just chicken scratch.
Two weeks later, I’d be demoralized because I couldn’t remember which conversation was which. More importantly, I was trying to give the team feedback on how excited people were about the product, but my notes fell flat. I’d have a great conversation, then two weeks later, my synthesis of 50 conversations would be six bullet points, and people would say, “Okay, cool.” We were missing so much in translation. We did a little prototype where we built a video playlist by downloading recordings and editing them together. The reaction was so much better-no more playing telephone. People could see what I saw: the excitement, the frustration. That was one input.
The other input was my background in engineering and product design. At UserVoice, I ran every team over about 10 years. Sometimes, as founders know, you end up running teams between leadership hires. I ran our sales team for about nine months, and my number one challenge was getting good notes. Some people gave too many, some not enough. No matter what notes I got, I always found myself asking, “Yeah, but what did they actually say? Can I just hear the 30-second clip of them saying they’re definitely going to buy or not buy?” I wanted to get past the sales team’s “happy ears.”
You put those two things together, and it was clear there should be a better solution. Companies like Gong existed, but after researching and talking to about 100 people using those products, I realized this was a much bigger problem than just sales-everyone hates taking notes. In 2020, transcription costs were about $3 an hour, which was expensive for the average user, but I believed the cost would go to zero. There were already several good providers, and AI was getting better. At the time, some investors didn’t believe in AI, but I thought it would improve rapidly. We also believed no one wanted just a transcript, but that it would be the foundation for more useful things.
We had a thesis: if you wait until transcription is free and really good, you’ll be two to three years too late to start this company. You need to do the hard stuff now-recording infrastructure, integrations, distribution channels. If we could be the first to give this product away for free when the cost dropped, it would spread virally.
Doug Camplejohn
(04:30)
That’s amazing. You mentioned Gong for sales folks, but what did the landscape look like when you got started? It’s incredibly crowded now.
Richard White
(04:42)
Right. Once AI got really good and transcription became free, it became everyone’s new “Hello World” app. But in 2020, when we started, there were only sales-specific tools-Gong, Chorus, and some “Gong Lite” competitors. Gong was $130 per seat per month, and even the knockoffs were $80 per seat. All expensive, all focused on sales.
From user research, we saw managers loved Gong because it helped them keep up with their teams, but individual contributors (ICs) were less enthusiastic. They didn’t want to wait an hour for a recording after a call, and the action items weren’t very accurate. If you’re a sales rep, you’re not waiting an hour to do your post-call work-you’re doing it as soon as the meeting ends. That led us to realize our product needed to be free and really fast.
Our first version took about half an hour after the meeting ended to deliver recordings and transcripts. I told the team it needed to be faster. We got it down to seven minutes, then five, then three, and eventually to 30 seconds. At 30 seconds, it felt almost instant, and that was good enough.
We found that immediacy was a killer feature if you want ICs to use it and to replace note-taking. Note-taking’s value is in its immediacy, so we had to match that.
Doug Camplejohn
Speaking of note-taking, the latest batch of tools-like Granola-are doing a hybrid of call recording and note-taking. What do you think about that approach?
Richard White
(07:21)
I think it’s a good interpolation between where the future is and where people are today. In the future, AI will write all the notes for you, but for now, there are people who like writing notes or at least having control. Augmenting their notes with AI is a great intermediate step, and it’s something we’re looking at too. Ultimately, I think AI will do it all, but maybe it learns from you over your first few calls.
Doug Camplejohn
(07:58)
I saw a review where you were compared against 21 other players and came out on top-congrats! You’ve crushed the G2 marketplace with reviews. How did your G2 strategy come about, and how do you ensure people review you?
Richard White
(08:31)
Thank you. We knew early on our go-to-market would be word of mouth. You bring this tool into meetings, and you have to explain why it’s there. It’s not enough for it to be free; it should be a premium brand-free and really good, both product and service. We built a support and success team before we had paying users. We wanted people to feel like they were getting thousands of dollars worth of service, even though it was free.
We engaged users a lot, even paying them for feedback after their first call with Fathom. Most companies don’t care about feedback, but if you offer $25, people believe you do. We spent a lot of time making the individual user experience really good and built feedback loops around early usage. That’s led to a staggering number of G2 reviews-we’re about to hit 5,000, with a perfect 5.0 rating and the highest satisfaction product across all of G2. It comes down to a really good product and really good service.
Richard White
(10:14)
You mentioned the bake-off. It’s fun building products in the age of AI. It used to be that your roadmap was whatever you wanted. Now, it’s almost like R&D-testing the latest models to see if they make our output 5% or 10% better. It’s easy to get AI to say something, but really hard to get it to say the right thing. There’s a craftsmanship to it. I’m proud of those bake-offs because they reflect our AI team’s hard work, hand-reviewing hundreds of outputs. There’s no substitute for that right now. It’s a brave new world for AI product competition.
Doug Camplejohn
(11:24)
One thing that stood out in the bake-off was the evaluation of summaries. You have an internal team manually reviewing outputs. Is there a customer feedback loop as well?
Richard White
(11:50)
There is, but it’s not as strong as our internal one. Customers won’t sit down and review hundreds of summaries. It’s more like a Michelin star chef experience-we taste in the kitchen before bringing it to the table.
Doug Camplejohn
(12:16)
That’s fascinating. We do similar things with our TLDR summaries. So much of it is about taste and prompt engineering, not just the models. Sometimes you get a summary back and think, “That’s better than I could have done myself.”
Richard White
(12:56)
Exactly. Now, the state of the art is using multiple models for almost every feature-sometimes from different providers. We care about speed, cost, and quality, so there’s a multi-step process to generate notes, using cheaper models first and more expensive ones later. It’s much more complex now. Prompt engineering was the focus 18 months ago; now it’s about orchestrating different models. It’s a fun way to build software, working backwards from what’s possible.
Doug Camplejohn
(13:49)
I’ve always enjoyed the design aspect of product development. This feels equally creative-before, it was just screens and data, but now it’s much more.
Richard White
(14:07)
Yes, it used to be everything was just a CRUD interface-digital cabinets. Now, it’s more like playing Jenga, testing which model solves which use case. The worst thing you can do is force a model to do something it’s not good at, because a new model might come out and do it better with two lines of code.
Doug Camplejohn
(14:53)
Do you ever roll out features and realize you need to go back and adjust because you got it wrong?
Richard White
(15:00)
It’s less about getting it wrong and more about realizing it can be a lot better now. We have a high bar for accuracy and quality before launching features, which means we’re rarely the first to launch something. We were late to launch AI action items because early models weren’t accurate enough. But six to nine months later, you can get something 20% better, which is amazing if your baseline is already 90–95%.
Doug Camplejohn
(15:50)
You’ve talked about the importance of choosing your go-to-market model and aligning it with product development. You started with a bottoms-up PLG approach and gave away the product for a couple of years. How does that align now that you’re in more mid-market and enterprise deals? How do you balance free users versus paid users?
Richard White
(16:20)
I’m a big fan of free to start. Monetizing is hard, but getting people to use your product is even harder. For a social product like ours, if it becomes a daily habit, that’s a high bar. If I can get you to use it every day, I’m confident I can get someone to pay eventually. We focused on the single-player experience first-retention, then onboarding, then virality. Now, it’s a more complicated chessboard, building features for our team workspace product, which is where we’re putting a lot of AI effort this year.
We still improve the free product as AI gets better and costs come down. We forced the market to offer free versions, but ours is truly free forever. We’re happy if you stay on the free version forever. Our model is to give value to individuals, knowing that eventually a manager will want to organize their team and pay. That separation lets us keep improving the free product while focusing on enterprise features.
Richard White
(18:41)
A year ago, the state of the art was AI writing good notes for your meetings. Now, AI can look across all your team’s meetings and pull out important moments, write internal docs, and more.
Doug Camplejohn
(19:03)
How do you think about the competitive landscape?
Richard White
(19:07)
There are a handful of competitors. When we started, the opportunity was clear because no one was doing what we wanted outside of sales. Many startups avoided competing with Gong, but I saw Gong as the ideal competitor-successful, expensive, focused elsewhere, and serving managers, not ICs.
I love competing against sales-led tools with PLG products. The real battle was building a viral engine. We’re now at “peak note taker”-like when Zendesk came out, and suddenly there were 30 competitors, then a few years later only two or three were left. I think we’re at that point now; only a handful have escape velocity. I expect a lot of consolidation, but it’s good for consumers and keeps everyone on their toes.
Doug Camplejohn
(21:32)
You touched on privacy. When you started, there was fear about recording calls. Now, every VC call has a tool running in the background. What are your thoughts on privacy?
Richard White
(22:02)
There have been client-side recording tools before, but we purposely avoided that and always used a bot in the meeting. It’s important that all parties know the meeting is being captured. Now, transcription is so good you may not need the recording, but I still think there’s value in seeing tone. Most of the data can be captured in transcripts, which are very accurate now. There’s a gray area legally with transcription versus recording, but we don’t want to support stealth recording. It’s a brand risk-someone could record their boss, leak it, and IT departments would ban the product. People are more comfortable with recording now, and even more so with transcripts. If it’s just generating notes, no one minds. Our tolerance for this has evolved, and I think that’s a good thing. It’s inevitable-like customer support calls being recorded.
Doug Camplejohn
(24:10)
People have gotten much more comfortable-or oblivious-about privacy. You’ve talked about your team being mostly engineers and success people. What’s the size of the company now, and can you share any metrics?
Richard White
(24:30)
We’re about 70 people today, across engineering, sales, and support. We just hired our first marketing folks in the last three months, which is kind of insane for where we are revenue-wise. Service really matters to me, so we’ve invested in high-touch support, sales, customer success, and engineering.
Doug Camplejohn
(24:57)
How would you define your culture, and how do you maintain it as you scale?
Richard White
(25:10)
It’s an interesting challenge as a fully remote company. We haven’t formally documented our culture, but we focus on principles rather than values. Low ego is one of our top principles. We hire really senior people who are also humble, which is a rare combination. Otherwise, our culture is what you’d expect from a startup-resourceful, high work ethic. I want us to be a $100 million revenue company with 150 employees or less, which I think is achievable with AI. That’s important for maintaining agility as we grow. Culture often breaks down when companies get too large, so we focus on hiring the right people and keeping the team small.
Doug Camplejohn
(27:01)
You ran UserVoice for over 10 years. What lessons have you brought over, and what are you doing differently?
Richard White
(27:14)
UserVoice was around for almost 12 years. I ran every team at some point. It was my “finishing school.” We started as a PLG company before that term existed and ended up as an enterprise company. It was a good, not great product, which forced us to get good at execution. That made me a T-shaped entrepreneur. When we started Fathom, I knew enough about marketing that we didn’t need a marketing person for four years. Starting Fathom felt like speed-running a video game compared to the first time. What felt like open-ended questions at UserVoice are now multiple-choice questions at Fathom, and that’s been super helpful.
Doug Camplejohn
(28:43)
You mentioned the joy of building and creating value. As a CEO, what keeps you motivated on tough days?
Richard White
(29:00)
Honestly, positive feedback-like G2 reviews-keeps me motivated. A previous co-founder once said I didn’t care enough about revenue, and it’s true; I care more about hearing people say they love our product. That’s what drives word-of-mouth virality. We have an internal “Boom” channel for every positive mention, support interaction, or review. That’s my lifeblood. Of course, we don’t always get it right, but external validation is a huge motivator for me.
Doug Camplejohn
(30:02)
If you could automate one task forever with AI, what would it be?
Richard White
(30:15)
I’m really excited for AI to come into the physical world. A good friend is building a robot company that uses AI models to clean up around your house and do chores. I have a notoriously bad memory, so I’m excited for AI tools that act as a second brain. But what I really want is a robot that fixes things around the house-like Judy from the Jetsons.
Doug Camplejohn
(31:08)
What are some of the most common questions people ask with Fathom’s “ask anything” AI interface?
Richard White
(31:25)
Most people use AskFathom for specific meetings, as a better search tool-like, “Did we talk about this?” It’s helpful for people who weren’t on the call. It’s also used to generate artifacts, like follow-up emails. Where I’m excited is going beyond single meetings. We’re testing a version that looks across all your meetings, so you can ask, “Show me every time someone had a positive reaction to our new AI feature,” or “Send me every negative reaction to pricing.” We even had it write a five-page article on the history of Fathom’s transcription backends by scanning all our engineering meetings. That’s going to be huge for knowledge management and ambient awareness.
Doug Camplejohn
(33:07)
What are some of the most surprising uses of Fathom you’ve seen?
Richard White
(33:10)
There are a lot of agencies, therapists, even churches using it. My favorite is agencies using it to keep clients honest-if a client disputes something, they can show the meeting clip. One agency said Fathom saved them $10,000 by settling a dispute with a client.
Doug Camplejohn
(33:59)
What’s something you’re passionate about that might surprise people?
Richard White
(34:05)
I’m a huge skier-I love to ski. Maybe that’s not surprising to people who know me. As a kid, I built with Legos and did startups in high school. Despite spending a lot of time indoors at computers, I love getting out on a mountain.
Doug Camplejohn
(34:39)
Where are your favorite places to ski?
Richard White
(34:42)
That’s a complicated question-it depends on the weather, snow, time of year, and who I’m with. Jackson Hole has a special place in my heart. I also love Vail and Beaver Creek, where I learned to ski. Even though I’m in San Francisco, I usually fly to ski.
Doug Camplejohn
(35:08)
My son is in Denver, and we’ve enjoyed exploring smaller mountains like Copper.
Richard White
(35:18)
Off the beaten trail, I’ve been to Revelstoke once and can’t wait to go back. Telluride is amazing. I have something good to say about almost every mountain, so it’s a tough question.
Doug Camplejohn
(35:31)
We’ll have to do some skiing next season. What’s one thing you can’t live without in your daily routine?
Richard White
(35:41)
Honestly, I have a meal delivery service called Thistle-amazing gluten-free, dairy-free, clean food. I’m so much more productive than when I made my own food or ate junk that made me sleepy. Getting a good lunch between eight hours of meetings is crucial for me not to be a grumpy, crappy version of myself.
Doug Camplejohn
(36:22)
Makes sense. Finally, how can listeners keep in touch with you and help Fathom on your mission?
Richard White
(36:29)
If you haven’t tried it, go to fathom.video and sign up. It’s completely free, free forever. If you have feedback or want to connect, find me on LinkedIn-it’s the only social media I use now.
Doug Camplejohn
(36:45)
Richard, thank you so much for your time. I really enjoyed the chat.
Richard White
(36:49)
Thanks for having me. This has been fun.
