Vital Learnings from Bootstrapping and Selling a $55k a Month Business with Arvid Kahl of FeedbackPanda

Courtland Allen: 7s

What’s up everyone? This is Courtland from IndieHackers.com, and you are listening to the Indie Hackers podcast. On this show, I talk to the founders of profitable internet businesses and I try to get a sense of what it’s like to be in their shoes.

How did they get to where they are today? How do they make decisions both at their companies and in their personal lives, and what, exactly, makes their businesses tick? The goal here, as always, is so that the rest of us can learn from their examples and go on to build our own successful internet businesses. Today, I’m talking to Arvid Kahl. Arvid, welcome to the show.

Arvid Kahl: 35s

Thanks for having me.

Courtland Allen: 37s

You and your girlfriend, Danielle, are the founders of a company called Feedback Panda and you recently posted an exciting milestone about it on Indie Hackers. I think you called the post, We Sold Our SaaS.

You explained that it took you two years to grow Feedback Panda from an idea into a fully-fledged business making $55,000.00 a month in revenue with the just the two of you, at which point you sold it. First of all, congratulations. That’s great to hear.

Arvid Kahl: 1m 1s

Yeah, thanks so much. It's been quite amazing.

Courtland Allen: 1m 4s

You posted this milestone a month ago now. Obviously, you were excited at the time. How are you feeling now?

Arvid Kahl: 1m 10s

Oh, I’m still excited. This is a thing that’s going to last us forever, because just the accomplishment alone is wonderful. This opens the doors to the people you get to talk to after this and the new avenues you can go to. It’s just great.

Courtland Allen: 1m 26s

Well you have one of those businesses that I think would be tough to sell. It’s just two of you, you, and Danielle. You bootstrapped the company so you both owned together 100% of the company. You grew rapidly to $55,000.00 a month in revenue in just two years. How do you decide to sell a company like that?

Arvid Kahl: 1m 43s

It took us by surprise. We never wanted to sell the company, but we built it with it being sellable in mind. I’ve been reading a lot prior to building this particular company, because there was a time when I was not the founder, when I was employed at a company in Hamburg in Germany while living in Berlin, which is two had a half hours away by train.

I was commuting three days a week from Berlin to Hamburg and back, which is five hours a day and fifteen hours a week that you sit in this metal box and you just zoom through Germany. If there’s one thing about Germany, it’s that the connectivity between cities is really bad. You didn’t have the internet. You couldn’t do anything, so what you could do is reading or listening to podcasts or audiobooks.

I was doing this for two years, so two years, three days a week of two and half hours of commuting, so I would listen to all the podcasts I could find, particularly this one. I think I went through the backlog of every single episode that you made.

I was reading a lot of books, among which was Built to Sell by John Warrillow. That really stuck with me. If you ever build a company, I was thinking - and I wasn’t planning to. I was just an employee at that point. If I ever build a company, I’m going to build this ready to sell at any point. Automate as much as you can. Take yourself out of the business, that kind of stuff.

That’s stuck with me. We built the company from the beginning like this, so that there would be optimizations, automations, all towards making it easy to exit, but we never wanted to exit. We just wanted to grow the company from the idea that it was in the beginning to helping as many teachers as possible with this. Feedback Panda was the SaaS for online English teachers. That’s why it took us by surprise to sell the company.

Courtland Allen: 3m 28s

What does it mean to build your business with it being sellable in mind? What are some of the tips you got from that book?

Arvid Kahl: 3m 33s

Well I think Warrillow describes an agency. It’s just a story in his book where he describes how somebody owns an agency, wants to sell it and suddenly notices, “Hmm. I can’t really sell my company cause I’m the company.”

If you have an agency and you do design work and these things, then you have to do the work. It’s like being a freelancer, essentially, in a corporate box, but it is still just you and your services. So if you are out of the company, if you sell it to somebody else then there’s no service in there, cause you were the company.

He describes for companies who want to be able to sell themselves at some point to somebody. You’ve got to just make it that you’re not needed. Build the company with you leaving the company in mind, or you not needing to be working in the company.

That means hiring people. That means building processes. That means building automations. It’s the opposite of what Seth Godin was talking about in Linchpin. The most important thing? Make yourself the least important thing, the most easily replaceable thing.

Courtland Allen: 4m 28s

Make yourself easily replaceable. Terrible advice if you are an employee but great advice if you're’ a founder trying to sell your company. I think it could be very tough to give advice to other founders selling businesses, because quite frankly everybody’s situation is different.

Oftentimes people will email me when they’re considering selling their company or they’re going through an acquisition, because I’ve been through the same thing before with Indie Hackers. Situations are just so different. If I look at your situation and mine, I was a solo founder. You have a cofounder. I was only making six or seven grand a month when I sold Indie Hackers. You were making $55,000.00 a month.

I went to work for my acquirer, Stripe, and I’ve been there two and a half times longer than I worked on Indie Hackers by myself, whereas you quit working on Feedback Panda after you sold it. So there are a lot of variables and it’s hard to give advice because anything from your experience might not apply to the person that you’re talking to.

Arvid Kahl: 5m 18s

It’s always the same, particularly with bootstrap businesses or indie hacker businesses. All advice is anecdotal. Everything you’re saying that you experienced in your business is depending on so many factors that are unique to your own business.

But in all this anecdotal advice is a lot of truth that is applicable to every single business, because there are things that work because they’re the right thing to do, and there are things that work because they’re the right thing to do for your unique business at the time.

You have to figure out which is which. A lot of advice that bootstrappers give each other is coming with the caveat that, “This might not be the right thing for you, but think about it. Think how this could be applicable to your own business and if that’s the right thing for you.”

Courtland Allen: 5m 58s

Totally. You said that your acquisition came as a surprise, that you and Danielle were happy growing Feedback Panda the way you were. Walk me through the story there. What happened to make this acquisition a reality?

Arvid Kahl: 6m 9s

We got an email. The story goes way back in a way. I think I told about this in the exit interview that we gave for SureSwift Capital, who we sold our company to. They had an exit interview with us and I could tell the story where it all started. And it started, funny enough, with an Indie Hackers podcast episode for me. You were interviewing Moritz Dausinger if you remember that, way back, who sold his two businesses.

Courtland Allen: 6m 35s

Yeah, Docparcer, Mailparcer.

Arvid Kahl: 6m 37s

Exactly. He sold those two to SureSwift. I guess it must have been three or four years ago. It’s been a while at least. I distinctly remember, we started our company, it must have been June or July 2017, around that time. It’s always kind of hard because in your mind you started way earlier and then you formalize into an actual company.

I remember the day I went to the bank to open a bank account for our company. I was listening to that particular episode, with Moritz. On my way back from opening the bank account, I listened to the last part of it, and he was mentioning that he sold his company. He sold it to SureSwift Capital.

I was thinking, “Oh, that’s interesting. There’s a company that would buy companies like this and that would have such a positive vibe with the founder who sold it to them that he would then talk about it? That’s cool.” So I googled them and forgot about them.

Two years later, we get an email from Kevin McArdle, who is the CEO of SureSwift Capital. I see the email and I immediately thought, “Oh, yeah, these guys.” Because I remembered back then your episode with Moritz, and it just looked full circle. The day we opened the bank account was also the day we stopped needing a bank account for the company, because it started the whole acquisition process.

So we exchanged a lot of emails. We got talking. We, Danielle and I, started talking to other people who sold to SureSwift, did our own due diligence on the company, on the company that would then acquire us. They, of course, did their due diligence on us.

That happened, then we got to an agreement and transitioned the company out. It was all extremely easy and extremely painless, but that was because we had all these optimizations and all these automations and all this making yourself sellable in place.

Also, I’m a German, so I keep tight records of everything, because we’re taught to do this in a way. And we had every document in place. I did a lot of documentation just because I like it, which I guess is kind of masochistic if you think about it. But documentation, to me, doesn’t just mean documenting a code, it’s also documenting processes and building systems and then making it easy for somebody else to transition into them.

All that allowed us to make it easy to transition the company from us to SureSwift. We had a good time. We still have a great time. We talked to them. It was wonderful. I met Kevin for the first time at MicroConf Europe in Dubrovnik, just a couple weeks ago. So we did all this without seeing each other, but it worked out well because there was a connection from the beginning, and of course the story goes way back.

Courtland Allen: 8m 55s

Yeah. That’s crazy to hear that. I’m glad the Indie Hackers podcast could play a role in at least you recognizing who Kevin and SureSwift were.

Arvid Kahl: 9m 3s

You’re responsible for it, if you think about it.

Courtland Allen: 9m 6s

Where’s my cut, Arvid?

Arvid Kahl: 9m 8s

Yeah, they’re coming.

Courtland Allen: 9m 11s

So what are you up to now that you’ve sold your company? A lot of people compare their company to their baby. I guess in a way, selling your company’s like selling your baby. What do you do after that?

Arvid Kahl: 9m 22s

Well the first thing that Danielle and I did was to take a vacation, because that’s the one thing that we couldn’t do in running a company, running a bootstrap company, a SaaS, that had global customers, for two years. We were working 24/7 every single day, from the first day.

From the first day we had paying customers, we were always there for customer support. We always had to maintain the integrations that we had. We always had to fix bugs when they happened and help people out and react to the market, react to our customers’ needs, all these things.

We never really were able to take a vacation. So that’s the first thing we did a couple weeks ago. We went to South Africa, finally, for 10 days. That was a lot of fun being surrounded by wild animals for a change and not just computers.

That was really cool, got us an opportunity to relax a little bit and to center ourselves again. Because if you work on a business for two years or more, I guess, without a break you have blinders. You stare at whatever you’ve been doing and you want to repeat and improve and grow and these things, but you never stop and look into yourself. You don’t reflect where you are, what your goals are.

The goal that we set for Feedback Panda, funny enough, that I set for myself in the beginning, was to get to $50k MRR. That was the biggest thing I could imagine. It was like, “Oh, my. That would be the most wonderful thing to build a business that does $50k in MRR.”

And then we hit that, and then we didn’t have any goals anymore We didn’t set them. We reached it and, “Oh, yeah. That’s cool. Let’s just continue.” That at least was my goal at that point.

What we’re going to be doing now is do some reflecting, do some relaxing for a change, and then just see what happens. Because having built, having run, I guess, and sold the company, allows us to help other people do the exact same thing, both in a consulting way, I would guess, and informing people, telling people.

I’ve started a blog called thebootstrappedfounder.com, on which I try to distill whatever I learned and all these things I learned into articles that can help other bootstrap founders get to where I got and further, and just see a perspective into a successful business that worked and why it worked, and how they can apply it to themselves. So that’s what I’m doing. Danielle is also finding herself again, and then we’ll do the exact same thing, find a meaningful way to help people.

Courtland Allen: 11m 46s

It’s interesting how, when you’re working on a company, it’s like going for a long swim or something. You’re underwater the whole time, and things come to an end and you finally come up for air. The whole world is different. You’re a different person. The times are different and you’ve missed a lot of it because you’ve been so single-mindedly focused on what you’ve been working on. I think it’s refreshing to get to the end of that process.

Arvid Kahl: 12m 8s

Yeah. The kind of self-reflection you can do when you're out of this is very intense, because you notice, “I had all these anxieties. I had all these fears all the time, and they were driving me into doing this, into not doing that,” all these things that I now can retroactively figure out about myself. I never really had the time to while I was working on the company.

That’s the thing about a bootstrap founder. We were two people. That’s already splitting everything in half, but if you are a solopreneur, you are both responsible for keeping stuff going and then responding to people who complain about stuff not going and fixing it, all these things at the same time.

Of course you have a hard time figuring out where you are, because you’re always wearing all these hats and in between all these seats. So now, finally, I have some time to figure stuff out. That’s a big benefit of not having to work on the business anymore.

Courtland Allen: 12m 54s

Let’s talk about the business itself. I want to dig into the early stories and how you started this. Oftentimes when I talk to somebody who sold their company, we’re going back five years ago, ten years ago, but since you did this so quickly, we only have to go back two years for you to talk about the beginning of your company. Why did you start Feedback Panda in the first place?

Arvid Kahl: 13m 11s

Danielle is both my cofounder and my partner, my life partner, and we were living together in Berlin in a small apartment. I was working half remote, half present in a software developer job. She, being a trained opera singer, was singing in Berlin. But as art is a very seasonal thing, she needed to make some money on the side, so she was teaching English online.

The way she was teaching English online was for Chinese companies that would hire mostly American teachers or people who are able to speak the native language, and they would then teach Chinese children English as a second language over the internet. That’s what she was doing, and she was doing that quite a bit because we had student loans to pay, so she was working ten hour days.

These Chinese schools are all essentially the same. China is very regimented there, and you would teach a student one-on-one for 25 minutes, and then you would have 5 minutes to take a break, get some coffee or do whatever, and then you would teach the next one. And that would happen all day. Ten hours of working is essentially 20 students that you would need to teach.

After that, the parents of the student expect some sort of feedback. That’s where the name Feedback Panda came from, because that feedback making, that feedback writing process, took forever. If you teach for 10 hours a day and you have to write student feedback for every single lesson that you taught, which is 20, and you take that 5 to 20 minutes to type that out, like what did they do?

Did they do well? Should they be doing 10 minutes to type out some stuff? What did they do? What did they do well? What should they be doing in preparation for the next lesson? What was being taught and did they understand? You take 5 to 10 minutes for each of these things. That’s two hours a day in additional work, and that was not being paid. They were only paid for the time they taught.

The schools expected them to do it in the five minutes between lessons but you can’t really do that. You can’t spend 5 to 10 minutes in 5 minutes and still have a bio break or get some coffee or whatever. It just doesn’t work. So Danielle was working two extra hours a day, which after 10 hours of teaching is 12 hours. We essentially didn’t ever see each other.

At some point she built her own little system to fix this problem. She started writing templates to then use for the repeat lessons that she would teach, because she would teach the same lessons over and over, just to different students, so the content would be the same. She would have a work document here and an Excel sheet there and all these things.

At some point she asked me, “Can’t we do something about this? Can’t we build something here?” Being a software engineer, the answer is always yes, of course. If you see some system that could be automated, as a software engineering you just jump at it, like “Oh, yeah. I don’t want to do it twice. I’m going to automate it right away.”

So under her guidance, because she knew exactly what she needed, I built a prototype. She also knew exactly what the market is, because at that point there were fifteen, twenty thousand teachers just like her doing the exact same job, having the exact same problem. This is not randomly guessing who might have a problem and what it might be. This was exactly the same problem that everybody had.

So we knew, if we figure out a solution to this because we knew it was the most painful problem that teachers had, they would really, really like it. And two hours of writing feedback, unpaid, turning that into what is at this point five minutes of using Feedback Panda, that’s substantial, because every single day that’s saving you two hours. If you are an online English teacher, you could just teach for two hours more. You could just make more money.

So at that point it was ten bucks a month, a subscription as a SaaS, teaching just one more lesson would already pay for the month’s worth of saving your time. So that’s where the product came from. We did some marketing, and by marketing I mean one Facebook comment, like a comment on Facebook on a thread, that was in one of these groups that these teachers have, and it just snowballed from there. It was word of mouth all the way.

Courtland Allen: 17m 2s

I love so many things about how you started, because it’s so different than some of the common mistakes that indie hackers make when they get started. I think the most common thing is, “Oh, I’ve got a great idea for this product or this service. I’m going to build it. I’m going to code it. And then I’m going to figure out who’s going to use this and what problem it’s going to solve and why they’re going to use it.”

You did it the other way around. You started off by identifying a problem, like, “Hey, my girlfriend is working 12 hours a day. Maybe it would be nice if she only had to work 10 or 11 hours a day or she could get paid more, and she’s doing the same thing over and over again.”

You also understood not just what the problem was but who had this problem. You had a very clear profile that it’s these people who are teaching for these Chinese companies and having to do these lessons plans. So you had a problem and you had a market, and then you worked backwards with somebody who was feeling that problem, Danielle namely, to figure out a solution. So I think it’s not shocking that you’ve had so much success with Feedback Panda when you started it off in such a great way.

Arvid Kahl: 17m 58s

It was a revelation to me, in retrospect, understanding what we did, because we didn’t really do it consciously as like, “Oh, we’re going to follow these steps and then we’re going to be successful.” It just happened to be a perfect situation, where we understood exactly who we’re going to sell something to, and from there on we figured out, what is the best way of solving their biggest problem?

Now in retrospect that I think about it, that is a great way of getting a product that at least has some sort of relevation (ph) built in. What you said, most people who post their great product ideas on Product Hunt are essentially hunting for an audience. They’re hunting for an audience that may have a problem that could potentially be solved by the project or product that they built, which, like you said, is the wrong way.

You want to find the amount of people that can reliably sustain a business just with their biggest problem that you then solve in a way that makes them want to pay for it. There are very interesting concepts. I’ve been reading a book by April Dunford recently about product positioning, and she was also talking about competition in these things.

If people have system in place that is self-built with Excel or with Word or just notes somewhere that they manually write, that is already validation. You don’t have to have a competitor in the space, an actual company that is like you. Somebody writing down notes on a notepad is already validation for at least a problem or at least the potential of the solution to a problem.

That is very interesting because I’ve since then looked into a lot of my friends and family and seen where they solve problems that way. There’s a lot of potential to build good services on what people already do manually. But it still has to be there biggest problem. I think that’s a qualifier that you have to put in there.

It’s not just any problem that they solve with notepads. People still write notes when they do a shopping list and they don’t use their phone because that’s just not that important. They don’t want to get their phone out and it’s not their biggest problem. I think being able to pay rent is more important than installing another app.

But if you find an audience that’s big enough to sustain you and find their most painful problem, then you’re halfway there essentially. And only then do you want to go and look into the solution to the problem, because you just want to reduce it back to the audience and then go on its problem solution.

Courtland Allen: 20m 17s

Exactly. And the audience that you’re building for, the market that you’re targeting, that’s the hardest thing to change. It’s the thing that you have the least control over. If you build the wrong product, you can change that. If you have the wrong business model or pricing plans, you can change those.

If your distribution channels aren’t working out, you can change those. If your team is not right you can hire new people, but if the market that you’re targeting doesn’t exist, or if the problem you’re trying to solve isn’t a real problem, then there’s really nothing you can do about that. You can’t conjure a bunch of people out of thin air who have the problem you’re trying to solve.

So that’s the weak link in the chain, and that’s where you should start, I think, if you're going to start a business. Do you remember the early conversations that you and Danielle had around this, around what the product was going to look like and how it was going to address the needs of other teachers?

Arvid Kahl: 20m 59s

It was like, “How much can we solve with software here?” Because it is still teaching. It’s still something that is between a teacher and a student. It’s a very relational thing. You have a student you want to build a relationship with. The student wants to be respected and also taught sensibly by a teacher. So you can’t just automate everything, because computers don’t have a soul at that point and they can’t be relational in that way.

We were always trying to figure out how much of the work that the teacher needs to do is tedious and automatable? That was a lot of the early conversations, because it was all about, like, “We can do a lot with templating, but we can’t template everything. We can’t automatically generate everything for the teacher.” And that shouldn’t be how a teacher then talks to the parent, either. It’s still about being personal and being accurate, correctly telling people what happened with the lesson.

But the thing is, the content would never change, so having a template for the content was perfectly fine. The templating system behind Feedback Panda was in the beginning the most important thing, and it is to this day the most important thing because not only did we build a templating engine, but we built a collaborative templating engine, because we figured out teachers really like sharing information.

It’s not a big surprise I guess, but teachers really love helping each other. It’s one of these industries where people are not envious of each other’s success, because if you’re a good teacher, then you are a teacher to another teacher. You’re inspiring.

That made our market quite easy, too, because we triggered people’s willingness to share good information about how they became a better teacher, and then got them to refer people to Feedback Panda, build up on the referral system that we then also built in the future. But people were already doing this without incentivizing them, because they love sharing information.

We built that into our system as well. We made our templating system collaborative. We allowed people to share templates with other teachers, them to be able to import templates and all these things, building a network effect into the product which then also exploded growth.

Courtland Allen: 23m 0s

This is good stuff. I hope people are listening closely.

Arvid Kahl: 23m 3s

Yeah. We thought, “How can we make this a usable solution to the problems these people have, that the teachers are so desperately fighting every single day?” Then the ideas just came. Danielle has been amazing at understanding teachers, being a teacher herself. She’s been amazing at designing a product, because I am not a designer. I’m a good developer, I guess, but I’m terrible at good UX, if I can phrase it like that.

But she’s been amazing. She’s been building a tribe of teachers around herself and around the product as well, allowing us to do marketing that is almost hands off because once you give information to a community that is willing to share, it gets disseminated automatically. You don’t have to trigger people. You just have to provide, and they will share the good news about the product and about how it helped them.

So while I have always been the technical part, making it technically possible, all the ideation, all the product design came through Danielle, because she was the person who had the problem. It was her product to dog food, I guess. Because she needed the solution, and she also knew what solution she would need.

Courtland Allen: 24m 12s

I love how much we’re talking about the market and the audience. We might be getting a little bit repetitive here but I just want to keep talking about it, because it’s so important and it’s so overlooked I think, especially if you’re the stereotypical founder.

You’re focused on the thing that you're building. What’s the going to look like? How are you going to build it? What tools are you going to use? What’s it going to be called? What are all the bells and whistles and features you’re going to put into it? That’s what gets you excited.

I think you need that creative energy. If that gets you excited, that’s great. It’s probably something that should be celebrated, because even having that creative energy is rare. But also, you should probably take some time to do some research and sanity-check your idea before you jump into the product, because the market is so key to your success.

Like you’re saying, the fact that you're targeting teachers, and teachers are a group of people who talk to each other frequently, and who aren’t super envious and competitive and who will therefor share the new tools and tricks and tactics that they learn. That allows for you to grow through word of mouth, if you build a product for teachers.

That’s not true for every market, and it’s also something that you can figure out before you even get started building your product, if you sit down for 5 or 10 minutes and think about, “Who is my customer going to be?”

So I think it’s a shame that so many people will jump in and start building without asking questions, like “Who is this for? What does these people like? Where do they hang out? What do they read and what problems do they have? How do they learn new things and share information? What do they pay for? What problems do they find valuable?” There are so many questions you can ask before you even get started, to figure out if your company has a chance of succeeding.

Arvid Kahl: 25m 40s

I guess it’s a big problem with software development in general, because I guess most people who are indie hackers, at least the hacker part of them, they’re software engineers. If you’re classically trained through a university, they don’t teach you anything about business. They don’t teach you anything about people to begin with.

If you’re coming from a self-trained place or if you go through a bootcamp, it still is focused on the technical skill. The thing is, if you have this tool, it’s like the nail-hammer thing. You have the hammer, everything looks like a nail, and if building a product is what your hammer is, then whatever you see, the first thing you jump to is, “I’m going to build a product.”

Then you think, “Oh, I’m going to use this kind of software this time, cause that’s cool. I’m using this platform or I’m using these services,” and you start building it. You scaffold your thing. You build an API instead of reflecting on the business side because you’re never taught.

As a software engineer, if you come from that background, you just don’t know about these things until somebody tells you. Oh yeah, and you want to make money, right? And it’s not just integrating Stripe, although that’s a big part, but getting people to come to your product, these things you never learn, unless you're actively engaged in the community, on Indie Hackers, on a diverse range of platforms like Twitter or Medium, if you’re actively looking for this information.

But that already presupposes that you know that you need to know these things, and nobody ever tells you. So I think that’s the problem. We just look at everything through a product lens, which is why there are so many products looking for a market.

Product/market fit to me is a weird, phrase, because you shouldn’t fit a product into a market. You should see a market and go figure out a problem. It is market/problem fit or something like that. The product is a result. It’s not something that is on the same level as the market or an audience.

Courtland Allen: 27m 28s

Yeah. I couldn’t agree more. I think you’re similar to me, in that you like to research and analyze things and plan ahead and try to prevent making certain mistakes. A lot of people are the opposite. They’d rather fire from the hip. They’re big on intuition.

My personal opinion is that if you’re a founder, you don’t want to leave things to luck. If you decide there’s an entire area of your business that you don’t want to think about, that doesn’t mean that it doesn’t matter. It just means that you’ve left it to luck. You might as well have hired luck as your cofounder and given him an entire set of responsibilities while you go off and do this other fun stuff.

For me, I used to be more like that until I failed at a bunch of businesses. I looked back and it was like, “Wow, these were entirely preventable and if I’d just asked a few questions I probably wouldn’t have made these mistakes.” So I wonder what it was like for you.

Arvid Kahl: 28m 13s

That’s exactly what I was thinking, because I also started and failed a lot of businesses before. Feedback Panda, to me, has been an overnight success many, many years in the making. I started my software career here in Berlin at a small software agency, worked while I was doing my IT studies that I never finished. I dropped out at some point.

But while I was doing that, I was working on type of three(ph) backend module kind of thing. Back then PHP was still extremely bad. It has improved, or so I hear. That was my first job. Then I stopped studying this. I went to a different university and studied political science, because I wanted to do something completely different.

At some point I got a tweet by a company in Silicon Valley in San Francisco. They had looked at my LinkedIn(ph) profile and thought it was interesting, to see if I wouldn’t want to work for them. A couple of weeks later I’m in San Francisco meeting these people. I got a job there, working remotely from Germany for them as well and worked for a VC-funded company and saw how that worked. That was also very insightful, to see what can happen at scale if there’s a lot of money, but also what does happen if there’s a lot of money.

We’ve been having this big discussion on Twitter about 1Password and the VC money. You see a lot of positions there, some very positive, some extremely negative looking at what we can see money can do. That was also very interesting for me back then.

I stopped working for that company at some point, went back to Berlin and started a company, another bootstrapped company with a friend that didn’t go anywhere because we didn’t do any marketing. Then I started another bootstrap company with another friend, that didn’t go anywhere because we didn’t have any payment system integrated into our marketplace platform.

So I’ve been part of a lot of projects that didn’t go anywhere because we were not prepared. Once you have an experience, you have your takeaways from it. I’m an avid reader. I read a lot. I read whenever I can and for that reason it was great that I had two years of five hours, three days a week of just reading and taking in information.

If you just take in enough of the correct kind of information it sticks and it changes your way of thinking about things. So I also come from a lot of failures, but they all have their little golden nugget of information and insight.

Courtland Allen: 30m 33s

Let’s talk about how you grew Feedback Panda early on, because that’s something that a lot of founders struggle with. How do you get that first paying customer in the door? Once you get them how do you get a second and so on and so forth. What was your game plan for finding customers in the early days?

Arvid Kahl: 30m 46s

Our hope was in the beginning that we wouldn’t ever force it on anybody, because we knew that even though it is a community of sharing a lot of information, and that is really good, if you do the wrong thing, they share that, too. So we were careful never to push our solution onto people.

We never did a, “Oh, this is our product,” post into any of these Facebook groups. And Facebook groups they were, because that is where a lot of the people who turn out to be online English teachers that work from home are. Most of them are female. Most of them are from the Southern states in the United States, and there’s a lot of stay at home parents there.

They, from the age group and who they are, they hang out on Facebook. So there were a lot of Facebook groups. In these groups, we really, really carefully put the link to our product into comments. We never put it in posts. We just responded to people who already were interested in feedback and how other people dealt with it.

So our strategy was to really slowly, really carefully talk to the right people in the location that was essentially their water cooler, where they hung out to chat about their work. We didn’t have an acquisition strategy other than let’s see if people bite. And they did bite, and they started sharing it, and we just amplified their voices as well.

We responded to their questions. We responded to their comments. We allowed them to communicate with us, particularly with Danielle. She wrote blog posts about the origin of Feedback Panda. She wrote blog posts about teaching English online.

Our marketing strategy was always very social media focused. We started doing the VI Panda, which is this week’s most important teacher, where we would find a teacher that was interesting from our user base, and just interview them every single week. We would do that for the whole time.

There was this catalog of interesting teachers that come from all over the world that teach maybe from home, or they are in Thailand. They are expats somewhere and they teach from there. These stories were so engaging that people found their way into our subscriber base, I guess, through engaging content that they could relate to. So relatable content and being careful not to push stuff on people I guess would be what we did as a strategy.

The engagement happened because Danielle was already in these groups and had been there for the whole time she was teaching. She’d been working her job as an English teacher for I think three or four months before we had that epiphany that we could solve this with software. Leading up that, she was already part of all these communities.

She was a normal member, a participating member of these communities. It’s not that she would jump in a community that she was never in before and say, “Hey, look at this.” But she was already there. She already had been part of discussions. She already had communicated and shown that she loved teaching.

So people were, I guess, not expecting the market, but it just expected - word of mouth marketing, cause that’s how people share other things in these communities as well. They have webcams, to use, good microphones. So there was already some sort of exchange of products and services going on, so we could latch onto that and put our product in there, too, but in a way that wasn’t threatening or pushy.

Courtland Allen: 33m 43s

It’s fascinating to me how much of your thoughts around growing Feedback Panda were around things that you wanted to avoid doing. It wasn’t just, “We’ve got this plan and it’s great,” it’s also, “Here are the things we have to make sure absolutely to never do. We can’t force our product on anybody, because if we step on the wrong toes.” Then world of mouth becomes a double-edged sword and you don’t want to be on the wrong edge of that sword.

You were also cognizant of the fact that you had these Facebook groups that were communities. They were lively and vibrant, but every community has its own set of rules and customs and traditions and norms. If you come in and participate in a way that’s not authentic to how that community normally works, then people are going to know. You’re going to stick out like a sore thumb and your grow strategy’s not going to work. I think a lot of people can learn from what you did in taking the time to understand who you were targeting and how to best get in front of them.

Arvid Kahl: 34m 31s

To me, it’s important, being a software developer that was not a teacher, to also understand who we’re selling to. I wanted to get how these teachers worked, in a sense of how they work internally, and how they worked, how their workday was, how they did what they did.

Being able to be a part of the community or at least watching Danielle be a part of the community, gave myself a lot of insight into the psychology of our audience. Going back to what we said earlier, if you think audience first, the psychology of the audience is of utmost important, because that also determines how an audience solves their problems.

There are audiences where people think they know everything already in their job, and they would never use any different system because what they have in place is perfectly fine. You see this a lot in restaurants, where people don’t want to upgrade their technology, because they feel “This is perfect. That’s exactly what we need. If we get a new system then our waiters are going to be confused.”

Honestly, in Dubrovnik, we were having dinner at some point and there was a waiter with his android phone, and he had this app where he would need to take the order. It took five minutes for him to get the order of six people in, because it was so complicated and it was so weird for him to use the system, because it was brand new and he didn’t know how to add, like, ‘Oh yeah, potatoes instead of fries.” It was super complicated.

So people seem to forget that their audience has a certain way of doing things or at least has a way of approaching change. Knowing that, knowing how these people in a certain audience, our audience of teachers, would react to change, react to other people suggesting things that they didn’t know yet, allowed us to carefully word our marketing material and allowed us to focus on alleviating fears that people would have that it would sound too mechanical, that it would be machine generated.

Online English teachers are already somewhat technical but then there’s a big variety within the rest. There are people who know everything about computers, and then there’s people who you have to tell to turn it off and on again. It’s a variety of skill levels.

Once you figure out what these are, it makes creating content for these people in the audience much easier. It allows you to build automation systems that hit the people who need it the most at the right moment with the exact correct information and word it in a way so they understand it.

If you tell somebody to reinstall their browser extension, to us, “Okay. Going to go to browser extensions and I’m uninstalling it and installing it again.” But to some people who don’t even know what a browser is, how do you communicate that? That allowed us to build self-help systems that would speak to everybody on different levels in the same approachable way cause we knew what the lowest common denominator was.

So audience research to me is the most important thing in a business, because it will show you what problems there are. It will show you which are the most important problems. It will show you how people will react to you, how they will react to your product, how you have to engage with them. It all starts with the audience. That is, I think, one of the most important learnings that I have from Feedback Panda, because it made us so clear.

Courtland Allen: 37m 31s

I love how your knowledge of your audience was something that affected every part of your business. It helped you come up with your initial idea. It helped you decide on different features. It helped you decide on your marketing messaging and how you phrase things. It’s something that suffuses throughout your entire business.

I wonder if you ever saw anything, when you were researching your audience, that gave you pause, because so far it sounds like it was all sunshine and rainbows. It was all good news all the time. Was there anything about your audience that made it particularly challenging?

Arvid Kahl: 38m 0s

Yes, I think so. There’s a couple of things. There’s always volatility in a job like this. It was essentially like Uber drivers. You’re a contractor. The people that we sold to were contractors to these Chinese companies. They were not employees but contracted out, and these companies, Chinese companies, had gigantic hiring sprees where they would try to get as many people as possible.

I guess in the Chinese way of doing business, you want to be the biggest so everybody else starves in the market and then you get the whole market. So that was the approach that they were doing with hiring the contractors as well, so they were scaling up the hiring intensely which was great for us. Cause we started with 15,000 teachers that were out market, and two years later it was 70,000 or 80,000 teachers that were our market.

So the market was scaling for us as well which was cool, but there’s risk in that. There’s risk that in this predatory market, the company that you support or the companies that you support through integrations or you target as your customer base just evaporates. So that was always a risk that the market could just implode. China is very well known for imposing regulations that destroy whole parts of the economy. So while they did impose regulation, they didn’t restrict the companies that much, so it was good for Feedback Panda both in the past and in the future.

It’s funny, China imposed a regulation that limits the time of day that your child can take online English classes, because Chinese parents and the whole “our child has to be the very best of all children in China,” they were pushing their kids to go to school, after school tutoring, and they would stay up until 10 or 11 at night being taught English online.

There were a lot of kids that just fell asleep and I saw that through the video. They just work through. The day was over and they still were supposed to do English online courses. So they enforced the regulation that said after 8:00 p.m., no more teaching children online, which is interesting because that has an effect on the companies, Chinese English schools.

Now they need to shift their whole hours in the days towards the morning. Then teachers who are in the United States, in China when it’s evening it’s the early morning in the United States so you have the time zone difference as well. It’s perfect for us in Berlin in Europe, because it was a 9 to 5 day for us, where it was a 2 or 3 a.m. until 11 a.m. day for the Americans, and I guess 6 p.m. to 11 p.m. for the kids in China. So we were right in the middle in many, many ways, so that worked out for us.

But that was one fear that the market could just collapse. The second one that is also very important I think, particularly for a bootstrap business, the teachers don’t make much money. No matter where you go, teachers are always underpaid. They’re always overworked and unsupported, which is a great market, if you think about it from a business perspective. It’s a really bad market if you look at from an employment perspective.

That also means that it is very likely that some of the teachers could not afford ten bucks a month. They wouldn’t have Netflix because they couldn’t afford that. So that’s why we priced it quite low, even though it was a tool that would allow you to make 30 additional dollars every single day that you would teach, which for most people would be 20 days a month, so we could have priced it way higher to still provide value.

But we were careful not to over price it, because that would leave out the people who would need it most. We were always quite lenient when it came to credit cards bouncing or these things, because in the low income market, you want to be supportive. That also was important branding for our company, to be quite honest. We helped people when they needed help.

When the credit cards would be overdrawn, we gave them a lot of time to fix it and help them out, and that helped us with our branding as well, because it was just genuinely helpful. We just wanted to help these people. That’s why we built the product to begin with. But yeah, market shift and people not being able to afford it, but it worked out both ways.

Courtland Allen: 41m 56s

You have all these different puzzle pieces that fit together. You've got a market where people talk to each other so your reputation matters a lot. You also know that teachers don’t make much money, so you’re lenient, understanding, supportive and affordable, so people talk about that. Word gets around, and that makes your business an even better, friendlier option for people in your market.

There are also Chinese laws, regulations and culture that play into things, where I guess one factor we haven’t talked about is how that affects the size of your market, the fact that there are these huge Chinese companies that are pushing to hire more and more teachers means that the number of customers that you can sell to is growing.

That means that your business is easier to grow. A lot of people start businesses in markets that are either stagnant or dying, and they wonder why it’s so hard for them to grow. They’re like, “Oh, I just want to capture 5 or 10% of the market,” but if the market’s only a few thousand customers, 5 or 10% of it’s not that much. In your case, you had many tens of thousands, or hundreds of thousands, or maybe even more teachers who you could sell to. I have to imagine that played a huge role in you being able to grow so quickly.

Arvid Kahl: 43m 0s

I think the important part here is that the market still hasn’t had a cap. It was both large enough and small enough to be good for a bootstrap business. Because the moment you have a gigantic market, you have these gigantic players, because obviously there’s a lot of money to be made.

But if you’re in a market that has, I guess at this point there’s 150,000 online English teachers, and this particular niche, online English as a second language for Chinese children from the age of 4 to 14, that would be maybe 200,000. It’s still growing, but there’s a limit, and that limit keeps the gigantic corporations away from you.

So for an indie founder, that’s amazing because you know there’s not going to be IBM throwing a product in there. It’s not worth it for them, but it certainly is worth it for you. So the audience, the market, being both large enough and small enough for a bootstrap business, that was luck for us, I guess. But we found it, we saw these properties, and then we knew we could do something with it.

So if you know what to look for, it makes it easier, at least to dismiss markets that are either way too small, just a couple thousand people. With a couple thousand people, if you have one or two people that are competing with you then all of a sudden nobody gets to make money because you’re outspending each other in Google Ads. Even marketing is expensive when you have to fight for these low volume terms.

And when it’s too big, then you have competition that you can’t fight. You might want to join them, I guess, at some point, hope to be bought, but I don't think bootstrap businesses start with that in mind. You don’t want to be bought by a big company, not by a gigantic company necessarily. You may want to exit, but maybe not that way.

I think you're connected with the audience of your market as a bootstrap founder. You don’t just do something because it makes money. If you’re going to throw yourself into this for years to come, you better do it for something you want to do and that you care about and connect with. That was true for Danielle and me with online English teachers because we saw their plight, her being one herself.

It was also true for the other projects I started before but failed at, to be honest. It was a local food startup and a startup for a photographer. I was interested in it, but not as much as this time, cause this time I knew exactly that there was a problem that we could solve, and not just a product looking for a problem, looking for an audience.

Courtland Allen: 45m 13s

You mention something that I think is interesting, which is that you want to be in this Goldilocks zone where the market you’re targeting is not too big, not too small. It’s just the right size, where you can make money and be profitable without having to worry about some huge company coming in and getting interested cause there’s too much money and they just want to take all of it.

If you look at products that target all teachers everywhere, you have these huge companies like Blackboard, which are just massive and would be very difficult to compete against. But you had a niche.

You were only targeting English speaking teachers who worked for this limited number of Chinese companies teaching English to these Chinese students online. Your market was the right size for you. If someone else listening is an indie hacker in their targeting market, how can they know whether or not their market is the right size?

Arvid Kahl: 45m 56s

There’s two answers to that. For a market that has no competition, like in our case, you try to figure out how many people have this problem, essentially. It was fairly easy for Danielle, because she knew exactly from her research she did in all the online communities of teachers, that everybody who would teach a certain amount would run into this problem.

There was a class of people that had a shared problem and would have a shared pain. For a market that has no competition, that’s the only way you can do it. Because the moment there’s competition, you can figure out, “What size is my competition?” If the competition is a gigantic corporation, then the market may be too big for you. If your competition is a couple of scrappy bootstrap companies, you might be in an interesting niche.

That would be my suggestion here, because we’ve also targeted a way too big market that had venture funded competition in the local food startup that I was talking about. At the time that we were trying to go for that, there was just a company that was being funded by this big German venture fund.

They rolled us out of the market in a way. We never got in there cause they had everything. They had connections to everybody and they were doing everything. So it was too big for us. Didn’t know that before. Should have checked.

But now I would just say, we look at the competitors and again, in April Dunford’s words, it’s not competitors just as companies, but also competing alternatives, the solutions that people use. If there is an Airtable solution that people use that deals with 100% of their problem, it might also not be a perfect market, because there is a solution in place that already works perfectly.

But if you see that people even struggle with their self-built solutions, I think that’s a good indicator. And you have to a lot of research. I think I have a post on my blog about how to determine the size of your SaaS market. I just released it last week.

I just try to go through B to B, B to C and B to BC, which is our market, business to business consumers, which is individual teachers that are both end consumer but also businesses. It’s freelancers and contractors, these people. There are a number of ways you can deal with that.

You can find the information about the market. I list a couple things, going to conferences or figuring out how big conferences are for B to B, or trade publications. They still exist. People still read paper magazines. Just figuring out a couple about these things will allow you to understand if the market is the right size for you.

B to C is always complicated because there’s a lot of people that could be a potential customer, but a lot of people that will not ever be interested in your product if you have the whole lifecycle, where people are located, early adopters and these things. So you never know.

It’s hard to figure it out. There are some ways of finding that information but it’s more complicated. But the moment you know who has the problem, and that these people are already looking for a solution, that is the moment you want to act.

Courtland Allen: 48m 50s

Let’s talk about operationally how you and Danielle were able to run Feedback Panda. Most people I talk to who grow a business of this size have a full team helping them out, yet you and Danielle never hired anybody fulltime to help you out. How were you able to make that work?

Arvid Kahl: 49m 5s

Well I think I should have hired. That’s one of the things I noticed in retrospect. I should have hired somebody for customer service way earlier, because we were always doing the customer service together in our company from the beginning, Danielle and I.

We’d just take all the incoming things on Intercom, that’s what we used, and all the messages that would come through the live chat and we would respond to it. Then we would build a knowledge base article. If anybody ever ran into the problem again we’d post it to our knowledge base so it would be automatically suggested. We were building an automated system, but we would still deal with every single incoming piece of feedback that people have, questions and these things.

So we dealt with that from the beginning. We dealt with that the first day and the last day, but it turns out if you have 5,000 or more customers, the volume of incoming information could be kept under control, but the time when it happens does not.

If you have 20, 30 people reaching out to a team of two every single day, they’re prone to be interrupting you at some point. They will interrupt you while you’re doing development work, and they interrupted me all the time. I was trying to build features and couldn’t because there was always somebody asking, “How do I delete a student?”

I would have to respond to them and then we’d go back, and being interrupted while you’re doing software work, you’re going to need half an hour to get back in, 34 minutes later, somebody else would send another question.

So it kept me from being a software developer. That’s the thing I like. That’s the thing I love to do. I love helping people too, but in the end writing code is what I love. So we didn’t hire. We managed to deal with it but we should have hired much earlier.

I was responsible, like I said earlier, for all the technical things, and in the end most of the customer service after 16, 17 months turned out to be technical cause we solved all the non-technical things through automation, through suggesting articles with videos on how to do stuff.

So all the things that would come through that would not be solved by the articles were technical questions. “Can you merge my data? I don't know how to log in,” these kinds of things where you need to have somebody who can look into an SQL database somewhere.

So I would do that and the development work. Danielle was always responsible for marketing, for leading the company, for leading the tribe of teachers around it, for doing the content work, for doing the product design, the product management. That’s how we split it.

Courtland Allen: 51m 16s

You mentioned that you were having trouble getting work done as a software engineer, and you were the only developer working on Feedback Panda. How did you get around that? Is that a problem that you were ever able to solve, or was it always inefficient the whole time, with you doing customer support and software development?

Arvid Kahl: 51m 30s

It forced me to do as much automation as I could possibly build. I had to build every single part of the software stack to be completely automated. That means deployments are automated, failovers are automated. Alerting and restarting the system needed to be automated.

A quick look into the tech stack, I built everything into Elixir, on the Elixir Phoenix platform, put that in Docker containers and threw that on the Google Cloud on a Cubanese (ph). That is the whole stack of Feedback Panda. We used Vue.js in the front end and a couple added turn-based apps for integrations in the browser extension, but the whole core of the product was an Elixir Phoenix cluster that was running on the cloud.

Every single step, on every single level, there was automation. If something would break, it would automatically come back up. If there were errors they would be automatically reported. It could not deal with stuff manually anymore, which is great. Turns out that having a lot of automation makes your company extremely sellable, cause you don’t need to be there anymore.

So I think we made a lot of mistakes. I think we made mistakes every single week, maybe every single day because that’s what being an entrepreneur is. You make mistakes and you figure out what it was and you try not to make the same mistake again. You do that a couple thousand times and you succeed, I guess.

But all these mistakes added up to us building a solid, reliable, resilient system, that we could hand over with

in a couple minutes, once we did. That was what the whole handover phase when we sold the business was. “Here. Log into Google. It’s still running. Enjoy.” There’s not much you need to do if everything is automated.

Courtland Allen: 53m 4s

And not just automated but also documented, cause as you said earlier, you’d already written down everything someone needed to do to run your business.

Arvid Kahl: 53m 10s

Oh, written down is not it. I actually recorded an 11-hour session of me walking through my code base and talking to myself as if I was the next developer that would work on that code base, just explaining all the concepts.

The thing is, once we sold the company and we hit the transition phase, the first thing we did was hire our replacements. It was so fun to give my then-hired replacement just the links to the videos I did. “Here. Enjoy. This is your first couple days. Just listen to what this is.” Because I then didn’t need to do it.

It was available for all the developers that would work on it in the future, and it made it very clear to myself, as well, what the breaking points were, what things could potentially lead to trouble down the line were it to communicate that in a way that, if I hadn’t done that, there might be tripwires somewhere in the stack.

But I was able to explain everything in a way in this video series that I did. I don't know why I did it. I just felt like I really needed to document stuff. That was the best way of doing it. It was a lot of fun.

Courtland Allen: 54m 11s

The German in you, Arvid.

Arvid Kahl: 54m 11s

I guess. Yeah, must be. There’s certainly a lot of forced documenting. But it was always fun. It was always enjoyable for me. If you write code, you document it for yourself, your future self that doesn’t know what the hell you just tried to do in this code two years ago.

So you put that in, and if you look at it from a business perspective, you document it for the people that are going to run the company in the future. That might be yourself. That might be somebody else. I think if you're a bootstrap entrepreneur, you want to be able to hand over the reins of the company but not the ownership.

That, at least, should always be the goal. If you keep it, you benefit because you’ve built this. This is your accomplishment, your value that you provide, but other people can do the work because you might be off doing something else. You have another project or you just need a break. So having all of this in place made it sellable, made it easy to hand over. It was a blast.

Courtland Allen: 55m 1s

Why didn’t you ever hire any fulltime employees to help out with Feedback Panda before the acquisition?

Arvid Kahl: 55m 7s

You know what? I just didn’t know how to hire. That’s one of these things I was never taught in any of the jobs I did before. I think Danielle also never hired anybody. I was afraid to hire, because I didn’t know. How does this work? Do I talk to them? Do I have to do that interview thing? How do I know that they’re good?

I was scared of hiring, and so I never hired. It was a big mistake in retrospect. I guess the next project will not have that mistake included. Again, one of these things you learn. But I didn’t know how to hire.

Courtland Allen: 55m 37s

You know, I think there are things like this at every business. Danielle Gross, who I’ve had on the podcast before, wrote a blog post called Dread Tasks. My brother calls the ugh fields, because when you run into them it’s like, “Ugh.”

But it’s things that you don’t want to do as a founder, stuff that is hard to do or there is a lot of uncertainty. You’re not sure how to get started or you're afraid of it for some reason, or it’s just annoying so you just put it off, and you never do it or you do it really late. You procrastinate.

Sometimes that can kill a business. I had a friend a few years back who ended up doing a round of layoffs at his company, and he can trace it back to a decision to put off enterprise sales. He’s like, “Yeah, I didn’t do enterprise sales because it was scary and hard and we procrastinated. I can trace these layoffs all the way back to that decision two years before.”

At Indie Hackers, for example, I’m adding groups to the community forum. I know that growing these groups is going to be a ton of work. It’s going to be like starting a bunch of little new communities where I need to post every day. I’ve been procrastinating because it’s annoying and not that fun to work on. But I think in your case, you were able to compensate. You were able to survive because you didn’t need to hire. You could just automate everything, which is great.

Arvid Kahl: 56m 47s

I guess it was a result of that. It was as result of me being unable to hire that I thought, “Okay. Well I need to deal with this now.” Because if I had known how to hire, I would have thought, “Oh, I’m going to pay somebody a couple thousand bucks a month and they’re going to deal with this, and then they’re going to do it manually all the time,” and I would have forgotten about it.

But that would have added additional costs to the business, in a way. I’m happy I didn’t know. Now that I know, because I had to hire somebody after all, I know both things. I know that you can avoid it, but you also shouldn’t. So it’s a win-win, I guess, win-win-win for the business as well. The one thing about tasks like this, and everybody tells you that, don’t go for the low hanging fruit because everybody goes for the low-hanging fruit. Do for the most dreaded task because that’s the one thing that nobody will go for.

I tried that in many ways, but hiring was the one thing that didn’t want to do. I think I’m more like an introvert inside. I don’t want to interact with people too much, at least not in these confrontational situations and hiring is in some ways. It shouldn’t be, but it feels like it is.

“Okay, a new person. What do I do? I hope I do nothing wrong,” so I avoided it. I avoided all the person things, and hiring is a very important person thing. I never did it. I would now. That is the one scary thing I didn’t do.

We did a lot of scary things. We did a lot of little experiments. We increased our prices by 50% at one point. That worked out, but it was scary to do. Making stuff almost 50% more expensive for a customer base like ours was a risk.

We introduced a referral system at the same time so that was give and take at that point, so that worked out. But we had a lot of these scary moments with the product that we had no problem going into. But once people were involved, it was not for me.

Courtland Allen: 58m 34s

I guess one of the downsides to not hiring, even if you can automate everything, is that it erodes your work/life balance. One of the best things about hiring is you suddenly have this other person who can make decisions on their own. Sometimes they make better decisions than you would have and they can catch little things that you don’ have to worry about.

So you end up finding yourself with these giant chunks of free time. Whereas if you’re automating stuff, that takes a lot of time up front where you have to write that code to automate things. So I think it’s valuable to learn both, as you eventually did. What’s your advice for another introverted software engineer out there who’s also not had any experience hiring and who maybe needs to confront that choice at their company?

Arvid Kahl: 59m 11s

The biggest things is, most of the people in the software development community are like this. Both the people that are hiring and that are being hired are essentially the same kind of person, or often enough. I don’t want to generalize too much. But most people who are looking for a job may also be the same kind of introverts.

I think this is changing and I think communities are getting more diverse, obviously. But the thing is, if you feel like it’s a challenge, just think about how much of a challenge it would be for the person on the other side, the person that is trying to find a job, the person that is trying to get into your product. At least they may already be interested in being part of it.

I would tell my six-months-younger self to talk to people. Reach out and talk to people and try to find somebody who really loves these kinds of dreaded tasks. Because as much as I like talking to teachers through Intercom, through our chat system, it was not the job I loved.

The job I loved was software development. So that was the job I couldn’t do, because I had to do all the customer service with Danielle. But still, there as a lot. It impeded us both to get to where we wanted to be.

So try to find somebody who loves talking to people and loves helping and solving their problems when you’re looking for a customer service person. Jump over your own shadow and reach out. Just talk to them.

Mathematically it’s always clear that paying somebody a couple thousand bucks a month for doing a job is better than forcing yourself to do the same job at the same time you’re doing your own job, getting anxiety, becoming stressed, not being able to sleep anymore. Obviously you don’t want to do that. But you have to overcome this inhibition to hire. I guess that’s what I would say.

Courtland Allen: 1h 0m 53s

You mentioned another difficult thing that you and Danielle ended up doing was raising your prices, not by 10%, not by 20%, but by 50%. I’m looking at your pricing page now, and Feedback Panda costs $15.00 a month, so I assume it was $10.00 a month before that. You have a very price-sensitive audience. Tell me about how you did that and how it turned out.

Arvid Kahl: 1h 1m 13s

You know what, we started out with two monthly basic pricing plans, $5.00 a month and $10.00 a month. That was our initial pricing. It was even cheaper than now. We scrapped the $5.00 a month plan quickly.

Generally the first big move we made was to turn that off, because the people who choose a $5.00 a month plan, on average, not every single one, are quite likely to be very, very complicated. They reach out with a lot of customer support messages. They are very price sensitive, so they think they are owed every single bit of your attention.

We noticed that the people who would go for that plan would not go for the $10.00 a month, our almost unlimited plan. The $5.00 a month plan we had was limited to, I guess, 150 students that you could have in your system. Most teachers never reached that number anyway, because they may not teach as much, but most teachers said, “Okay. This is the professional plan. I’m going to pay $10.00 a month.” That’s the pricing psychology there.

But the people who bought the $5.00 a month plan and then complained to us all the time, we just couldn’t handle the workload. There was just so much going on. There were a lot of charge backs. There were a lot of people complaining about us not adding more features at the same time. I don’t want to sound too negative, but we had a lot of negative experiences there, so we scrapped that.

Then we were either the $10.00 a month plan or $110.00 a year plan. We had discounted yearly, which is also great. That is one of the best things we ever did was adding a yearly plan from the beginning, because the capital influx that comes from having a yearly plan where people pay $110.00 instead of $10.00 every single month adds up if you want to invest in better infrastructure, if you want to pay somebody to work for your company.

Having a couple dozen people who give you a couple thousand dollars ahead allows you to scale much more in a sustainable way. It allows you to do much more capital-intense stuff than if everybody is on a monthly recurring plan.

Also committing to a year is a pretty good indicator that you might be on to something with your product. If people think, “I’m in an almost temporary job,” because online teaching is temporary for most people, “and I’m still committing for a year of this product,” that is as pretty clear story at that point.

So after a year, we noticed that we added a lot of features to our product, but we still charged the same price. Not that we cut it off at the $10.00 a month plan. We had the cloud, Feedback Panda Cloud, which was a collaborative template sharing system, but we also had added a machine learning system to work in the background to do some fancy pronoun translation stuff. We had a text expander were you could add quick text manipulation things. We had built a product that was much better than the product that we sold for $10.00 a month.

It was clear to us that we should probably charge more. Then we did the thing that Patrick Campbell will never accept. We grandfathered all of our existing customers into the $10.00 plan. We should probably have done that on a limited time scale, grandfathered for a year and then increased them up to $15.00, but again, price sensitive customer segment, lots of goodwill, lots of good community around a product.

We didn’t just want to flip a switch, make $60,000, $70,000 a month instead of $50,000 or $45,000 or wherever we were at that point, but have people hate us for forcing them to pay more money. That didn’t vibe with our kind of audience. So we made the choice to grandfather everybody in.

We announced it a month in advance. We told them, “For everybody who subscribes after December 31st, the new price will be in effect but if you subscribe before, you get the old price.” We had a lot of spike around Christmas. That was cool.

And then we introduced the referral system, which to this day has been extremely successful. Teachers love referring other teachers for all different things. Forty percent of incoming signups come through the referral system, which is great. It says a lot about a working referral system right there, because we did it with a dual incentive in the beginning.

Courtland Allen: 1h 5m 14s

So would you say that that referral system is now the most significant thing you’ve done to grow Feedback Panda, or have other things contributed more?

Arvid Kahl: 1h 5m 22s

You know, that is a very interesting question. I think the most important thing to contribute to the growth of Feedback Panda was the fact that we built a collaborative tool that had people talking and sharing with other people built in.

I think that is the main driver of growth, because it’s the network effect in action. Everything you add to the product multiplies the value of the product. It’s not just an addition. It’s a multiplication. The referral system has been instrumental to sustained growth so yes, it is very important.

The great thing is teachers like ours already knew how referrals worked because they work for companies that have referral systems built into their own system. So it’s not just teachers referring Feedback Panda to other teachers, they get paid by the schools to refer new teachers, the friends and family, your old friend from high school.

“Do you want to teach English online for Chinese kids? Here’s my referral link.” And then they get onboarded and they make a couple hundred bucks for a referral. So they knew exactly that there’s value in referring. It was super easy for us to build it in. That was as low-hanging fruit, to be honest. We did it a year in. We should have done it much earlier, but it works. I think building a network-based system is very important to growth.

Courtland Allen: 1h 6m 32s

We talked about this a little bit a few weeks back. We’re both fans of this book called Hooked, written by Nir Eyal, who’s also come on the podcast twice to talk about his books, including Hooked. I use Hooked to determine how I build Indie Hackers. It’s informed a lot of my product decisions.

But you also used it to help you build Feedback Panda. I know we’re pushing it on time here, but I would be remiss to let you go without diving in a little bit on how you did this, cause I think it’s so helpful for founders, especially, who are building apps where there’s some social component, some sharing component, to think about this. So how have you used the model from Hooked to build Feedback Panda and make it collaborative?

Arvid Kahl: 1h 7m 11s

The Hooked cycle has been instrumental. I’m grateful to Nir for thinking of that. In the last part of the Hooked cycle there’s a trigger action variable rewards, and the last part is investment. At some point I figured out that investment, in our case, means putting a template that you would use as a teacher into your own database on Feedback Panda.

By having that, then being shared to other teachers, it would trigger them and it would allow them to import the template and it would allow them to get a new template that they wouldn’t have to write for themselves. They would use it, and when they wrote a new template, they would invest it into the platform and it would go back and trigger an action, variable rewards, an investment.

The cycle was an integral part of our collaboration tool. I guess it’s also an integral part of every referral system. Every referral system has that built in, too. But we made it a core of the engine of the product. That has just led network-based growth from the beginning, from when we built it in.

So I’m extremely thankful for Hooked. It’s an amazing book, just explaining how we can build habit-forming products for good. It’s not just necessarily getting people hooked on your product. We wanted to build a product that people would use every single day and would benefit from every single day. You don’t just want to drain money from their bank accounts and not have to deal with it. We wanted to build something really meaningful that enabled them to make more money, have more free time, spend more time with their kids. Hooked is the number one book that I recommend to every single founder, because the mental model of the Hooked cycle is extremely valuable.

Courtland Allen: 1h 8m 41s

It’s crazy how deep you can dive into every aspect of your business. If it’s the product, you can read books like Hooked. There are years and years of courses and classes and trainings and books you can read and take to get better at building products.

If it’s the market, there’s a whole bunch of stuff on market research and talking to customers. If it’s the business model, you mentioned Patrick Campbell. He’s got a ton of information out there about monetization, how much you charge. Distribution channels, don’t even get me started. It’s just an endless array of information.

I like what you said about Hooked and about the fact that you can build a habit-forming product that helps people develop good habits. You did that with Feedback Panda in helping teachers, and that’s what I’m working on Indie Hackers as well, trying to get people to develop habits that help them build better businesses.

Arvid Kahl: 1h 9m 23s

Yes. It’s about enabling people. If you can habitualize a good thing, like going to the gym or reading a book or writing in your journal every single day, that changes life for the better. These kind of things, if you habitualize them, they make you a better person. If you can build a product that has habitualization built in and provides value every single time people use it, that to me is a net positive.

Courtland Allen: 1h 9m 47s

Well listen, Arvid, you have a wealth of experience, not just from growing and selling Feedback Panda, but also from starting businesses that didn’t work out and from reading a ton of books. What would your advice be for the average indie hacker out there who’s just listening in and wondering whether or not they should start a business, too?

Arvid Kahl: 1h 10m 3s

Well I think everybody should try. Everybody should start something that they really, really care about. I think building a business has never been as easy as today. You don’t need to be a developer to build a software business.

You just had an episode on no-code. You can build a viable business without touching a single line of code. You don’t need to. Eventually you might want to, at least that’s me the software engineering talking. You might want to get into the specifics at some point. That is also just a perspective. It doesn’t have to happen.

If you have an idea or at least if you have an understanding of an audience and their problem and then have an idea, that is important to just enable and help other people. That’s I think where all business should come from. Of course it has benefits who run the business, who own it. We sold our company. That’s great. It’s a life-changing amount of money. But the most important part to me is that we helped tens of thousands of teachers do their job better.

If you want to do that, if you want to impact the life of other people, and if you want to impact the value that they can create by enabling them, then you should start a business. Then you build a product. Then you should find the problems and solve them, and make people get more capable than they were before. I think that’s the purpose of all business, all entrepreneurs should go for that. I would hope that everybody tries and starts the thing. Not everybody will succeed, but everybody could.

Courtland Allen: 1h 11m 31s

Everybody could succeed. Arvid Kahl, thanks so much for coming on the show. Can you tell listeners where they can go to learn more about Feedback Panda and the new blog and newsletter that you’re working on?

Arvid Kahl: 1h 11m 41s

Feedbackpanda.com is the website of the product. You can also find it on Twitter, at @feedbackpanda. You can find me on Twitter at @arvidkahl, which is A-R-V-I-D, K-A-H-L. That’s also my handle on Indie Hackers, so you can also find me there.

The blog is thebootstrappedfounder.com, where I also have a newsletter going on. I started the newsletter because I want to write every single week, so I want to hold myself accountable. That’s why every single week there’s going to be a post and there’s going to be an article and there’s going to be a newsletter with things from the bootstrap world that I found interesting, my learnings, all these things that I want to share.

Courtland Allen: 1h 12m 16s

Very cool. I’ll be subscribing. Thanks again, Arvid.

Arvid Kahl: 1h 12m 19s

Thanks so much.

Courtland Allen: 1h 12m 20s

Listeners, if you enjoyed this episode and hearing from Arvid, I would appreciate it if you gave him a shoutout. He is @arvidkahl on Twitter, and his website is thebootstrappedfounder.com. Feel free to subscribe to his newsletter and show him your support.

Also, if you're interested in receiving the newsletter for the Indie Hackers podcast, I send it out every Monday with each new episode. You get my thoughts on every episode, my takeaways, and what I thought was interesting. That’s at indiehackers.com/podcast. Thanks so much for listening, and I will see you next time.