Jack and Paul interviewed
Jack and Paul talk to Brian Casel about starting and growing Fathom Analytics.
January 24, 2022
When did Fathom begin?
In 2018 I (Paul) envisioned a simple + privacy-focused analytics product and shared a mockup on Twitter. From there, I found a technical co-founder who had to move on quickly after we built an open-source version of the product. Jack came on board in early 2019—then Fathom started to snowball in terms of our MRR and come into its own as a product.
How did you get your first wave of customers?
Initially, my audience was the primary source of customers to pay for Fathom. At the time, I was running a popular newsletter and had an active Twitter account.
As I mentioned, Fathom started as a tweet of an image of what I thought analytics could look like, and that tweet took off. From there, the first co-founder and I built out an open-source version, which millions of people have downloaded. From there, we launched a paid (hosted) version, Jack came on board, and we started to hone in on releasing features and making our infrastructure the fastest on the planet.
There was never really a giant wave or surge of customers at any specific time from any one thing we’ve done. Instead, it’s been a slow and steady ramp-up of customers and awareness for Fathom.
Are open-source and paying customers very different types of users?
Yes and no.
Managing their own servers, installing software, maintaining databases, and monitoring security is fun... for a small number of people. Those folks lean towards our open-source version of Fathom. And nowadays, the paid version is very different, has many more features, infrastructure improvements, and much better compliance.
For most people, even if they have the skill set to run the open-source version, it’s too much time and work, especially when we charge such fair prices to manage everything for them and serve both their script and dashboards as quickly as we do.
Most companies don’t want to hire a developer or team of developers to manage their analytics; they want it done for them, so many people quickly moved to our paid product after it launched.
A lot of the value proposition of Fathom is that we help collect, manage and serve up analytics data ridiculously quickly—while ensuring compliance to privacy laws.
Open-source also has a lot of challenges, like dealing with the deluge of issues, questions, and requests from people who don’t help contribute to a very popular project (or financially compensate the people who create and manage it). We decided early on to run Fathom in a way that best suited our lifestyle and goals, so we focus on the paid version of Fathom now.
Also, if “owning personal data” is why people think open-source is best, Fathom doesn’t store any personal data (it’s all anonymized), so there’s no real gain there.
Did your (Paul’s) audience have a higher churn rate in the early days since they could have been more interested in supporting you than solving the pain point Fathom solves?
There may have been a handful of initial folks who left because they wanted to kick the tires for my latest product (at the time, I was releasing a lot of new products), but it wasn’t noticeable in our churn rate.
For Fathom, the problem is pretty universal to businesses online: they all want some way to glean insight into what people do on their site and where they came from.
People will churn out if they don’t have an actual business, and paying for analytics doesn’t make sense for a website that doesn’t exist to generate or hopefully generate income. But for everyone else, $14/month is an almost unnoticeable expense.
Another thing is that our product becomes more valuable the longer a customer uses it. So if you’re a customer for a year, you can look at trends over that whole time (vs someone with a few days of data). So partly why our churn rate is so ridiculously low is that we have a product that continues to be more valuable the longer it’s used.
This is why we are one of the few analytics companies that offer unlimited data retention—meaning, you can view your analytics as far back as you have them. We’ll never delete or hide your data.
Our customers can also just “set it and forget,” meaning they don’t need to alter code or do anything to get value from Fathom past the initial setup. They can add our code snippet once and then look at our dashboard or get email reports for their required insights.
We’ve also invested heavily in compliance, so most of our competitors (even Google Analytics) can’t offer the same. That’s why we’ve got features like EU isolation, ad-blocker bypassing, and compliance without cookie notices.
Early on, we decided we couldn’t just rely on my audience to drive growth and awareness. So, it’s been “The Jack and Paul product called Fathom,” instead, it’s just “Fathom Analytics,” and Jack and Paul are just the people who run it. So we’ve worked on additional distribution channels and marketing ideas since the beginning.
What were some of those early marketing ideas you tried?
We’re also very value-focused on our content game, driven heavily by telling interesting stories from which other people (customer or not) can learn, especially on the engineering front. The article on when we got DDoS’ed is the most popular article from our blog to date.
Things like the Above Board podcast are a great content distribution channel we’ve used since the early days. There’s zero focus on “selling Fathom.” Instead, we focus on sharing our journey and what we’ve learned from it that others could benefit from knowing.
Our affiliate program is intentionally generous because it’s a great way people share Fathom with the people and businesses they’re connected to.
We had an article go viral on Hacker News early on, but it didn’t lead to people signing up for a Fathom trial. But the point was that it generated a lot of awareness, which led to trials and customers later on.
You also don’t need an audience to build awareness. It could be through SEO or doing podcast interviews or anything else. Since I had an audience, awareness made the most sense through that channel. Jack now has a mailing list of thousands to distribute engineering articles to, which has helped.
Our marketing plan has always simply been to: create value and awareness.
Speaking of marketing, did ProductHunt have any impact?
It helped, but we also had a brand new product, and I had an audience to promote our ProductHunt link. It didn’t make or break our product (nor does it with any product). Those surges or blips of interest can be lovely, but our focus is and will continue to be slowly generating awareness and value over time.
We also had several influential people in their areas try and love Fathom early on, which helped us become a trusted name in the analytics space.
These small, incremental events slowly compounded over time to move us from a few trials a month to many, many daily trials.
When did that switch flip that Fathom was a sustainable business that you wanted to sink all your time and efforts into?
Version 2 for sure. When that launched, we began seeing a massive uptick in signups. But not only that, signups shifted from people in my audience to people who just knew about Fathom from their business connections or contacts.
Then revenue grew to where Jack and I could pay ourselves actual salaries. That meant we could ramp down the other projects we were working on to focus almost solely on Fathom.
Were you (Paul) intentionally looking to build a business that was a way out of having an online audience? And since you (Paul) famously deleted your newsletter and left Twitter, did that impact Fathom?
It did just happen, but as it was happening, I realized it was my chance to stop doing something I didn’t like (being a “known person on the internet”) and move into something I did like (building and maintaining a fantastic product).
As soon as I realized I didn’t have to do what I was doing, I leaned heavily into Fathom being the brand, and my work became very much a secondary sort of thing.
Once I deleted my newsletter, personal site and Twitter, there was zero impact on Fathom’s MRR. Our growth has actually increased since then (but obviously not because I deleted my online presence, haha).
Do you see any less impact when you have something new to announce or people to hire?
We have fantastic reach with our company Twitter account, which continues to grow and has excellent engagement. We now get more traction from Fathom tweets than we ever did from @pjrvs tweets.
We have no plans to hire aggressively because our growth rate (MRR) is ever-increasing. We’re pretty conservative regarding spending (more on this next episode!) and will continue to be in the future. We don’t hire unless it’s too painful not to, and we don’t hire unless we can easily remain profitable when we factor in a new salary or retainer fee.
Sometimes adding more people can slow things down. And it is for us even, which is expected (as Jack is writing a lot of tests now, so the new person can just jump in and start confidently writing code).
How big is Fathom in terms of people at the moment? Just Jack and Paul?
Yes, but that’s changing!
We’ve got a small and trusted group of experts on retainer, like our Privacy Officer, our EU infrastructure expert, a content writer and an SEO expert. But we’re also bringing on a full-time developer in February.
We aim to keep things lean and minimal while questioning growth at all times, but we need help since Fathom has become popular.
How do you keep things lean as you grow?
Part of it is just the nature of the business: we don’t need a lot of people to build and maintain a company like Fathom (it’d be different if our business model were like Amazon’s).
Another part of it is how we built our business: bootstrapped and sustainably, so we don’t have investors who expect us to increase our headcount rapidly.
We also spend a great deal of time building features (so they’re easy to use) and how we document those features.
What are the things you currently want to grow with Fathom?
Because we’ve now existed for several years, a lot of what we’re working on currently is optimizing for the things we already know work well for us. Like content and story-telling, SEO, engineering articles and sharing.
We’re getting a content newsletter started to better distribute things we write or record with the folks interested in learning from them.
We are also reasonably aware of the specifics in achieving product-market fit and can double down on those things. Features like ad-blocker bypassing and EU isolation have been huge for us because we know our customers want those things (and, in turn, those features will drive more signups).
By listening to the people who pay for or are thinking about paying for Fathom, we can pay attention to trends and adjust our internal roadmap to solve those problems, leading to even more awareness.
There are no plans to do anything radically new or different than we’re doing. Instead, our objective is to continue to pay attention to what’s working and why it’s working. When we process this information, we just have to be reasonable and consider how each new feature or pivot can help the most significant number of people already using our product.
How do you handle the deluge of feature requests you get from running a popular product?
One of the biggest things we factor into decisions for what to build next in Fathom is: what will have the most significant impact on the greatest number of customers. Because we’re still in a growth phase, we want to add more to Fathom to make it even more valuable to those who pay for it. So we prioritize the features that all customers can benefit from and use regularly.
We also keep notes on everything everyone requests, so we can look for patterns in what people ask for (and how many people ask for each feature). That way, we aren’t catering to a single customer’s wishes and whims; we’re catering to what most customers want from our software.
We listen to every feature request from customers because they care enough to want Fathom to be better. And we feel like it can be.
You can listen to the full interview on our podcast, Above Board
You might also enjoy reading:
- Sharing MRR is egotistical
- From almost shutting down to tripling our MRR
- Making the world’s fastest analytics even faster