I did an Ask-me-Anything on ProductHunt this week, and one question stood out to me. Jon Jackson asked me about the kind of life I led while running the business, how stressful it was, and what that looked like.
Doing AmAs gives me a lot of opportunities to reflect on those things. Now that there are a few years between running (and eventually selling) the business and today, I’m able to look back from a slightly less subjective vantage point.
Building a SaaS was a gratifying activity at all stages. But over time, stress and anxiety crept into the day-to-day of running the business. Here is what happened during the two-year journey of FeedbackPanda, the EdTech productivity SaaS I co-founded with my partner Danielle Simpson.
It was a lot of fun in the beginning. I got to build a fully-fledged software product, and I had a designer, product manager, marketer, and tester in Danielle, who used the product for herself.
I only rarely felt stress then. It was more like having to solve little challenges and puzzles to advance the product. Since our marketing was word-of-mouth-based and we had a very tribal customer community, there was plenty of buzz around the product, and people eagerly tried it out.
In the early days, people didn’t depend on us that much, or at least we didn’t feel like they did. Our fledgling SaaS was a newcomer to an utterly underserved market where people had been scraping by with self-built spreadsheet solutions. We were a bonus, not a necessity.
Since we didn’t have that many customers — maybe a hundred or so a few months in — we could handle the support load. We made a point of turning any first-time answer to a problem into a knowledgebase article immediately. Over time, this allowed us to have repeat questions answered automatically with high-quality, step-by-step solutions.
The technical complexity of the FeedbackPanda product was relatively low in the beginning. There wasn’t much to fix on the product side, which was great for keeping stress levels down.
But that didn’t last long. We had made an infrastructure choice that came back to haunt us just a few months into having paying customers. We hosted the FeedbackPanda servers on a Docker-based hosting platform provided by a small German startup. We thought we’d support our local startup ecosystem and deployed our product with them. It ran fine for many weeks, but then something broke on their end.
All of a sudden, we experienced hour-long outages that our hosting provider struggled to recover from. We had no insight into what happened, our service just started failing randomly, and it got really bad. At one point, there was an almost 24-hour-long outage.
Not being able to do anything about this was extremely upsetting. We were still trusting that they’d fix their issues eventually but started doubting that when another week of daily outages went by.
Even worse, we were traveling from Germany to Canada at the time, and not being able to do any customer support during our flights was very frustrating.
As the outages got worse over time, we needed to find a way out. Eventually, with the help of a very dear friend, we scrambled to switch our backend over from the small hosting provider into the Google Cloud. We did that within a few hours, and it immediately stabilized the product. Tech choices are important, particularly in the beginning. We learned that the hard way.
With uptime being back to normal, we reached out to our customers and tried to apologize for all the issues. Almost all of our users gave us another chance, but I was left scarred a bit: I had developed “email notification PTSD” from the customer messages and the monitoring alerts. Whenever a notification sound would pop up, my heart started racing.
Even today, two years after selling the business, I still physically react to email notifications, and the sound of customer service chat windows opening. It’s getting better, but it’s not going away.
As we grew our customer base, we also expanded the scope of our product. We built integrations into the services our customers used on a daily basis. The problem here was that we had no means to know when the Chinese Online Teaching platforms that we integrated with would change their interfaces. They would constantly change their service at random times, and we’d need to react as fast as we could to adjust our integrations.
Since our integrations were browser extensions, that meant fixing the extension, deploying it to the Chrome and Firefox addon store, and hoping that the automated update would update them before our users noticed the extension being broken. It was a cat-and-mouse game that we couldn’t ever really win.
That resulted in high baseline anxiety for me. “Will they change it while I’m working? Or will it be in the middle of the night?” were typical thoughts that I carried with me every single day.
I tried building systems that could alleviate this tension. I scraped their update fileservers, built automated testing systems that ran continuously. Still, no matter what I built, these things could only help with the speed of discovery, not anticipate it enough to make me feel at ease.
At the same time, more and more customers started using our product. Any noticeable problem would now be reflected back at us in more customer service conversations.
Since we never hired anyone to help — Danielle and I had been running FeedbackPanda with just the two of us as founders — this often created another stressful kind of situation: whenever something happened, we needed to both solve the problem and communicate it being solved at the same time. Context-switching like that got very distracting very quickly.
My biggest mistake was not hiring anyone to help, either a customer service agent or another developer, to help me fix those issues. I thought I didn’t need to find help until I couldn’t keep up with the load anymore. But my mind tricked me: I thought I could still keep up with it even when I was extremely anxious and making mistakes because of it.
And because I was experiencing those situations more and more, I veered into a severe bout of burnout just over a year and a half after we started FeedbackPanda.
When we sold the business in 2019, a big load fell off my shoulders. In fact, my motivation to sell was primarily impacted by my need to get rid of the source of all that anxiety and stress. I would have very likely stuck around longer had it not been for my inability to overcome my hiring inhibitions. Danielle tried to encourage me to seek help, but a strange mix of indie founder pride and a complete disregard for my mental health made me ignore her very sensible advice.
Now you’ve heard my story, let me share a few things that I would recommend in hindsight.
- If you are overwhelmed, don’t dismiss this feeling. It’s an indicator that you need to either solve a problem or find someone to solve it for you.
- Hiring someone to help is not an entrepreneurial failure. It’s a sign of success. Suppose you are struggling to work on multiple things at the same time: delegate. You’ll make fewer mistakes and feel better.
- Keep an eye on the expectations you set for yourself. A growing business will stretch you out; that’s how scaling works. More customers, more features, more work. You can’t magically find more hours in the day. You won’t be as responsive a year into the business as you were on day one. And that’s okay.
- Listen to your peers and business partners. I thought I knew better. I should have listened to my co-founder, who had my best interests at heart. She saw what happened to me and suggested ways to remedy my issues that I couldn’t see clearly. Try stepping back from your subjective reasoning and invite in opinions of others.
- Build systems and processes that reduce workload. Whenever you can, automate and document. Automation will mean less time spent on tedious tasks, and documentation will make it easier to eventually give the job to someone else. Even before you hire, documentation will be great: if you can follow a set-by-step checklist, you’ll make fewer mistakes when push comes to shove.
Ultimately, you can’t avoid the fact that a growing business will also grow the sense of responsibility you’ll feel as a founder. There will always be stress, and there will always be anxiety. But you can keep them in check by knowing how they show their ugly heads.
Take a deep breath, and build a calm company that allows for a (mostly) stress-free workday.