AMAs January 2, 2020

I’m Arvid, a software engineer turned entrepreneur. I co-founded, ran, and sold The EdTech SaaS FeedbackPanda

Arvid Kahl @arvidkahl

I’m Arvid, a software engineer turned entrepreneur. I co-founded, ran, and sold The EdTech SaaS FeedbackPanda after bootstrapping the company for just two years. I have recently started a blog called The Bootstrapped Founder where I share my experiences from FeedbackPanda and all the less successful startups I’ve been part of before.

I have been building software products for over 15 years. I love building things, and I love helping other people do the same.

I'll be here on Wednesday 8th of January at 10am Eastern time to answer your questions.

Ask me anything!

  1. 6

    Hey Arvid, I really enjoyed your podcast. Very inspiring story! What is next for you? Any new projects in the pipeline?

    1. 1

      Thanks so much, Daryl! Great to hear you enjoyed the podcast. I had so much fun, and I think it's audible :)

      The great thing about selling a company is that it gives you room to breathe for a bit. That is also the scary part when you come from a life filled with work every day of the week. That was the most mentally taxing part of being acquired: coming to a complete halt after going as fast as I could for years.

      At this point, both Danielle and I are spending time with family, and we're looking into new opportunities. Figuring out what you want to do when you have so much choice is a whole new problem to deal with.

      For myself, I have started The Bootstrapped Founder blog with the intent to turn it into something big in the future. I have found that I love writing and helping other founders. Sending out a weekly newsletter and thus forcing myself to write at least one article a week allows me to write (better) and support other founders who want to build a sustainable bootstrapped business.

      The blog, the newsletter, and anything that is still in my overflowing list of thing to write is the next project for me. I really enjoy writing, so this will be my main activity in the next few months.

      On the technical side, I find myself coding little tools that support my weekly research into interesting topics and conversations that happen in the bootstrapper sphere. I have the feeling that there might be a SaaS-able solution somewhere in there, and that I might build that sooner or later, but I'll have to do much more market and problem verification before I start building anything serious here :)

  2. 5

    Hi Arvid, was the choice of Elixir/Phoenix a topic of discussion during the exit negotiations?

    1. 2

      While I can't talk about the specifics of the conversations we had there, you definitely won't find as many Elixir-based SaaS apps in the market as you would find Rails or PHP apps. For many acquirers, that poses the risk of not having the right talent to deal with maintaining the product after the founder who built it finishes their transition out of their business after selling it.

      There are a few ways of helping potential acquirers not to see this as a problem: bring them a developer who can deal with the stack as part of the company, or make it very easy for them to find one. In the Elixir world, there are a lot of highly competent devs waiting for an exciting Elixir job, and they all hang out on the Elixir slack or frequent the Elixir forums and job boards like Elixir Radar.

      And then there is the quality question: if your code is well-documented, tested, and reliably deployable, the amount of work to maintain it is significantly reduced and much less of a threat to the acquiring business. I went the extra mile in preparation for the sale and made an 11-hour video walkthrough through the codebase to make absolutely sure that even if a novice developer would look into the project, they could find their way.

      In conclusion: it makes a difference, for sure, but you can do a lot to make it less frightening and encourage potential acquirers to look into the wonderful technology that is Elixir/Phoenix.

  3. 4

    Hey Arvid! Would love to learn more about the 10+ years of software you were doing before FeedbackPanda! What did you enjoy doing most during that period, and how did it set you up for FeedbackPanda?

    Looking back, could you have "beelined" to FeedbackPanda?

    1. 2

      I started programming when I was in 9th grade. Turbo Pascal was all the rage, and I loved being able to coerce a computer to do things for me. Eventually, I graduated to Delphi and started building desktop applications. It was in school still that I (privately) learned about JavaScript and the web. In 2003, I enrolled at a university for Computer Science, which I dropped out of a few years later. But in the meantime, I had started a job at a web agency.

      And that is where I learned about scripting languages and backend code. I worked with PHP, building Typo3 extensions. It was both great and horrible. But it did set me up to understand that there are so many ways of dealing with problems. Frontend, backend, applications, scripts, I figured out that the language mattered very little as long as the problem got solved.

      That opened up a mental path to programming enlightenment. From then on, I took an interest in every language, trying to figure out what it would be good for. At some point, that spilled over into other technologies, such as databases and similar components.

      I built a two-sided marketplace with CoffeeScript (yes!) and MongoDB. I built a low-bandwidth-focussed asset distribution system with Node.js and Postgres. A few Ruby and Go projects happened along the way as well. I learned how to use Docker and Kubernetes Orchestration. I worked on a highly fault-tolerant IoT messaging platform with Elixir/Phoenix. I loved every single experience, as it was always something new, either technically or topically.

      And finally, FeedbackPanda happened, as a dockerized Elixir/Phoenix application running on Kubernetes, connecting to a Postgres database. It's the culmination of all prior projects.

      I don't think there is any beelining in entrepreneurship. I read "all the books" and I still had to learn so much. Knowledge can prepare you and keep you from making grave mistakes, but you will experience failure, and you can only develop experience by actually doing things.

      I have a mantra: It doesn't matter if you succeed or fail. If you work smart enough on the solution to a meaningful problem, you can have two outcomes: either you win big, you succeed, you arrive. Then, everything is great. Or, you can fail. Then, you are now a domain expert, having gained valuable insight, ready to teach others through consulting or transferring your experiences and knowledge into another industry. Either way, you come out of your journey ahead of where you started. Repeat until you succeed.

      1. 2

        (Apologies - been away from my computer for a while)

        I love this answer so so much. People gravitate towards stories of instant success - but more often than not, it takes some wandering, working on things you love, and many years for the fruits of your actions to compound into success.

        It's the culmination of all prior projects.

        Thank you for sharing this!

        1. 2

          You're very welcome. I think we get told that "the big success", the "moment in the sun" is a singular, noticeable moment in time, but it's really a gradual change, and it really only requires showing up and sticking with a thing for long enough.

          I like the idea of the opportunity surface that is growing over time. With every skill improvement, every experience, every mistake, you just ever so slightly increase the chance of good things happening. It's cumulative.

  4. 3

    Hey Arvid, I know I am late to the party, but... Did you have any hesitations on selling when you did, versus waiting a couple more months?

    I know many SaaS companies sell for 3-4x ARR, so that means for every additional $1k MRR you grow, you could potentially increase the sale by $36k-$48k.

    Was there any part of you that thought "let's just wait until we hit $60k MRR and sell for an additional ~$400k"?

    1. 2

      Ah yes, that was a big challenge for me. Every additional dollar you accomplish in MRR is worth a few dozen more when you sell. I had to think long and hard to come to the following conclusion:

      It's fine to say "this is enough". In a world of chasing every little ounce of growth, where "more" is always perceived as "better", you have to look at what matters to you, as a person. You will always be able to work just a bit longer, spend just a bit more time on things, and the rewards will be just a bit higher as well. It's what we learn to do. Grow, grow, grow. Never stop.

      To me, I felt it was time to stop chasing those numbers because I felt that I had accomplished what I had set out. When Danielle and I asked ourselves what would be an ambitious goal, we thought that 50k MRR would be amazing. Here we were, at 55k+, so that goal was accomplished, and we never defined any higher goal. So there was no goal we desperately needed to get to anymore. All I needed at that point were good, solid reasons to justify stopping.

      Back then, here is what helped me say "this is enough":

      • Our growth was good: We looked like a juicy acquisition target because we were one. We had solid numbers and the financial projections looked amazing. Selling a business is never easy, so having these figures made us look like a low-risk business, and that means good money

      • Future growth is not guaranteed: I would rather sell the company while the projection is shooting up to the top right. Any business experiences a plateau sooner or later, and we had not reached that point yet. A plateauing business means less money, a growing business means more.

      • Our methods experienced diminishing returns: We started to scale, so what worked with a few hundred customers turned out not to work so well with a few thousand. We didn't spend much time learning how to do things at that scale because we were caught up in the day-to-day, so we could have either hired someone to do it (and work through all the changes needed) or sell. Selling sounded like the less stressful option.

      • I wanted to reach financial stability sooner than later: While this is very personal, it was one of the most important reasons. I wanted to get into the post-economic state of mind as soon as I could. Operating from a place of security and comfort was a life-long goal, and it started to be quite possible if we sold at that point. Like I said, enough is enough, so that made the deal very enticing.

      I hope this shows you what I was thinking at that point. After the fact, you always think "what if", but I am extremely happy with the outcome, and I don't regret a thing.

      1. 2

        Man, really appreciate the well-thought-out response Arvid.

        Definitely makes sense.

  5. 3

    Hey Arvid, thanks for doing this!

    I emailed you this question couple of weeks ago, probably it slipped through the cracks:

    In your IH interview you talk about that first Facebook post introducing FeedbackPanda. Do you think the audience would have responded differently if you said you were building an app (=not ready yet, nothing to play with) instead of 'we've created an app, here go try it'?

    1. 1

      I think that people would have reacted differently.

      Particularly when talking to a non-technical or non-entrepreneurial audience, you will run into problems when you don't have something tangible to present. As a founder, you will have grand designs, wonderful ideas, and a wild imagination. Your prospects won't. They are likely burdened by lots of work, their own dreams and problems. You can't expect them to join your optimistic imagination and think of a product that doesn't exist.

      But you can infect them with your passion.

      If we hadn't had an actual product, we would likely have made a video. Danielle and I would have talked passionately about building a solution for teachers, mentioning all the problems we knew they have and how we would solve them. We would have show sketches and mockups of the application and asked for immediate feedback.

      But if you want to have people truly join, give them something that helps them immediately. Wait until you have something that helps, even if it's just a little bit. We waited until we had something to share, and only then did we talk about it.

  6. 3

    Thanks for the AMA, Arvid. Plenty of questions! (I have yet to hear the podcast episode so sorry if it has already been answered there.)

    I understood that the idea for the business came from your partner's job. Had you, as a developer, considered bootstrapping a software business before? If so, did you have any specific idea in mind?

    Did you build it entirely on the side, or did you quit your job at some point? It's not super clear from your story because you also mention that you automated a lot and made yourself "disposable" / "replaceable".

    Did your partner keep teaching with the support of the tool (like, "eat your own dogfoot")?

    Why have other startups you've been part of been "less successful"? Which mistakes did you identify and actively avoid with FeedbackPanda?

    You recently tweeted, "Founders often undervalue building and leading a tribe.", did you have one with the teachers? How did you build/grow it, while building the product?

    What motivated you to sell relatively early? I imagine you could have scaled this further on your own, from school to school, country to country?

    Checked the imprint on archive.org, seems like you've been the (legal) sole proprietor, so I assume you sold the assets/IP? How hard was the contract for the sale of assets, any issues w.r.t. German law vs. US law?

    1. 1

      Thanks, Matthias! That are quite a few questions :) I'll try to answer them all.

      I understood that the idea for the business came from your partner's job. Had you, as a developer, considered bootstrapping a software business before? If so, did you have any specific idea in mind?

      FeedbackPanda was not my first rodeo. I've been part of a few startups before: some failed outright, some still exist, some never went past the idea stage. In most cases, we had no outside financing, so bootstrapping was the only way to get started.

      I was well aware of how complicated it can be to build a product and a business at the same time, particularly when you're a solo founder. That's why it was such a breath of fresh air when I found such a wonderful partner in Danielle, who brought not only a great idea, but also domain expertise and knowledge of our audience and their troubles.

      That had always been a deterrent for me. I didn't know enough about people outside my little domain of software engineering. Of course, I had all those ideas about dev tools and building things for other engineers, but it's a sad fact that developers are not a good audience to start a SaaS business for. It's the DIY mentality that is a blessing and a curse: it allows us to build great tools, but then people won't pay for them as they'd enjoy building things themselves.

      I had always hoped to end up finding some constellation that would allow me to bootstrap a business where I had product, tech, and audience/problem covered. FeedbackPanda was exactly that.

      Did you build it entirely on the side, or did you quit your job at some point? It's not super clear from your story because you also mention that you automated a lot and made yourself "disposable" / "replaceable".

      Initially, I built the whole product on the side, effectively moonlighting. It was a weekend and work-night project, and the prototype was created fairly quickly. Because I really liked my job in the IoT field, I didn't want to devote all my time to the project. As we had paying customers very quickly, I needed to build a lot of automation just as fast to deal with customer support while both Danielle and I were working. I ended up quitting my full-time job for personal reasons, and after a while decided that instead of finding another job, both Danielle and I would go full-time on the business.

      Did your partner keep teaching with the support of the tool (like, "eat your own dogfoot")?

      Definitely! Danielle was teaching in the beginning, as I built the prototype for her, essentially. I wanted to be able to spend time with my girlfriend! And there she was, sitting and writing Feedback for hours. That wouldn't do! So when the prototype reduced her extra time spent on this tedious work significantly, we knew that this is something for other teachers as well. But the intitial phase was 100% dogfeeding, and it made the product work extremely well for the niche that we wanted to build it for.

      Why have other startups you've been part of been "less successful"? Which mistakes did you identify and actively avoid with FeedbackPanda?

      So many things we did wrong in the past!

      In one case, we had build a great product, but the person who had the product idea and was responsible for actually marketing and selling it into their professional network just stopped working on the business as it didn't grow fast enough for them. Nowadays, people would be much more aware of how long it might take and that you should just struggle through the "trough of sorrow", but back then in what I think was 2013 or 2014, we lacked that awareness. The startup fell apart.

      In another case, I partnered with a domain expert that wanted a solution, but didn't want to work on specifying it enough for me to actually build something. No market analysis, no prospecting, no customer interviews, no design prototypes. Just "let's build something". That never went anywhere.

      In general, all of the problems that led to startups not working out can be traced back to erroneous or skipped problem validation and verification. I recently wrote about finding the critical problem - and I mean it. If you build a product first but never check if it solves a problem that people think is painful enough to pay, you'll have a bad time. With FeedbackPanda, we made sure that we were solving the critical problem of our audience, and that we had an audience that was both big enough and small enough to allow us to build a bootstrapped business.

      You recently tweeted, "Founders often undervalue building and leading a tribe.", did you have one with the teachers? How did you build/grow it, while building the product?

      Danielle had been an active voice in the ESL online teacher community before we started the business. She had been speaking up about teacher's needs, and she was part of the important conversations. Once we released the first public version of FeedbackPanda, she started building a tribe around herself as a teacher. A teacher who cared about other teachers. A teacher who cared so much that she had built a tool to help other teachers. Consistently engaging the community, staying on top of the conversations and giving helpful advice and encouragement helped us grow a tight-knit community of teachers around FeedbackPanda. If you want to learn more about Tribes, I recommend reading the book "Tribes" by Seth Godin. Danielle did, and it sparked the idea for building the FeedbackPanda tribe around our product.

      What motivated you to sell relatively early? I imagine you could have scaled this further on your own, from school to school, country to country?

      I personally found myself to become more and more stressed with my work in the business. I was one of two people, and I had the development work as well as much of the customer service to deal with. We never hired (and that was a mistake), so we just tried to automate as much as possible. It still was a lot of interruptive stuff happening, so it bred anxiety and stress.

      At a certain size, our bootstrapper methods didn't work as effective anymore. Growth after 50k MRR is a different game than growth at 5k MRR. I personally felt I hit an imaginary skill ceiling. I didn't really know what to do, and unlike before, I didn't care to learn it. I wanted to build, to create, to invent. But the company neede reliable, repeatable processes to grow further. So I started looking for someone who would want to do that.

      Checked the imprint on archive.org, seems like you've been the (legal) sole proprietor, so I assume you sold the assets/IP? How hard was the contract for the sale of assets, any issues w.r.t. German law vs. US law?

      While I can't speak to the specifics of the sale, I can say that our acquirer was extremely professional and our lawyer called the contract "surprisingly boring." It was an extremely easy, smooth, and uncomplicated transaction, and we didn't run into legal issues at any point. In general, asset purchases are much easier to do internationally than share deals, even though they are more heavily taxed. But we had everything set up correctly too, so if you make sure the business is correctly strucutred and all financials and documents are in order, this is essentially a non-issue.

      1. 2

        Thank you so much, Arvid! I greatly appreciate the time and effort you've spent to answer in so much detail.

        1. 1

          Happy to share what I know. Thanks for asking such interesting questions! :)

  7. 3

    Just discovered your blog. How many articles are you planning on writing this year?

    1. 1

      Thanks! At a minimum, I will be writing one article a week. I begin every episode of the newsletter with a short summary of that article, so I have to write at least one!

      I recently began to write another series called The Emotional Journey of a Bootstrapped Founder, which I will release new articles for every few weeks.

      So, in 2020, I expect at least 52 articles, but it's going to be more like 80. Because I love writing. And I have taken a lot of notes :)

      1. 2

        Awesome, I read a couple of your articles and like the way you write. Hopefully I can finish the blog over the next week or two.

        1. 1

          Thanks so much for those kind words. I'll be working hard to continue providing articles worth reading :)

  8. 3

    Thanks for taking the questions. My question is how do you come up with good B2B ideas?

    1. 1

      I think the best way to come up with B2B ideas is by finding an interesting niche and digging deep into their problems. Find their critical problems, find the most critical one, think of a solution. Check that this solution actually helps them with their critical problem. THEN, and only then, start building a product.

      Let's say you're interested in tech recruiters. Start by talking to them about their job. Where are they professionally? Where to they want to be? What is keeping them from arriving there? How do they measure efficiency and effectiveness? Amount of emails sent? Amount of workers placed? Which one of those has the biggest impact on their success? What do they need to feel accomplished? What makes them more reputable? What helps them help themselves?

      Asking questions is how you come up with ideas. Ideas form after you take in the problems and issues that people in your target audience talk about. Problems first. Ideas later.

  9. 2

    Great talk!

    Thanks for sharing to much Arvid.

    I wonder if you could detail your strategy to setup your accounts and information to facilitate the possible due diligence in the future.

    Could you share your ideas behind this?

    1. 1

      Keep your business and your personal accounts separate. This starts with bank accounts. If you want to make sure you have a flawless separation between your private funds and the funds of your business, get a business bank account as soon as you can. While this will incur some fees in most cases, it is also good practice, and it will protect you from overdrafting your personal accounts in case of something going wrong with your business. Depending on what legal form your business takes, you will be required to have such a separation anyway.

      It's even more important to have your service accounts kept separate. Create a separate Google account for your business and use that to log into the services you use for your company via OAuth2 or using your business email and a strong password. Keep all of the login details in a separate 1Password vault, and keep only logins and secure notes related to your business in there. That way, when you hand over your business, all you need to do is to invite your acquirer into that vault, and all relevant credentials will be available to them immediately.

  10. 2

    Do you have a proven way to cold outreach to prospects. What is your email like?

    1. 2

      I'm sorry to say that I don't have any template for this. I didn't do cold outreach, as we went to where our prospects were: social media groups, mostly on Facebook. We became part of the community and did our outreach the social-media way: by commenting and engaging with relevant content. We became part of the community before we did outreach.

      Warm outreach is a lot of work. But we were a network-effect-reliant product, and we needed word-of-mouth to happen, as our customers are a tight-knit niche community, and it was much easier for a customer to convince a prospect than for us to do it.

  11. 2

    When you would start tomorrow a new business, how would you do it? (how to pick the right idea, build MVP etc.)

    1. 3

      The first thing I'd do would be researching the audience I am interested in, and then scheduling at least five video calls with prospects.

      1. Audience Verification: I would select prospects that show clear signs of CARING about what they do. People who are outspoken about their work, people who are proud to be who they are and what they do. Find the Innovators (and "early" Early Adopters).

      2. Problem Exploration/Validation and Audience Validation: I would prepare a list of questions all aimed at determining their critical problem and the size of the market I could expect. I'd ask the questions in the call but mostly listen to stories and complaints that come up naturally. Best record the call or take meticulous notes.

      3. Solution Exploration/Validation: After that, I'd start figuring out how I could leverage my technical knowledge to solve their problems. It's all about transferring technical ideas from one industry into another. I'd work closely with the people I called to talk about the usefulness of my ideas. Once I have found something that seems to excite people, I'd have my solution.

      4. MVP: Now, I would build a little prototype, maybe a NoCode or LowCode version, or a really simple CRUD-based prototype. I'd have a few of my prospects try it, and if they show signs of using it, iterate heavily to make it accessible for Early Adopters.

      Then, the business part starts. Pricing, Marketing Pages, Customer Support systems.

      Phew. I will write about all of this in great detail very soon.

  12. 2

    Now that you've had some time after the sale, what does your reading / podcast list look like? Have you taken up any new hobbies?

    1. 2

      Funny enough, not much has changed. I spend a lot of my time writing, and only occasionally read or listen to podcasts. While I attribute this to the fact that I am traveling a lot, it's also a shift in focus from consuming to creating.

      I listen to the Akimbo podcast by Seth Godin, Bootstrapped by Steve McLeod, Bootstrapping SaaS by Val Sopi, Bright & Early by Brian Rhea, Build your SaaS by the fine folks of Transistor.fm, Freakonomics Radio, The Indie Hackers Podcast (of course), the Ladybug Podcast by Kelly Vaughn, Emma Bostian, and Ali Spittel, Masters of Scale with Reid Hoffman, Out of Beta by Matt Wensing and Peter Suhm, Rogue Startups by Dave Rodenbaugh and Craig Hewitt, The SaaS podcast by Omer Khan, Slow & Steady by Brian Rhea and Benedikt Deicke, and many more.

      My reading list is overflowing, but I am currently working on Atomic Habits and a few other books that deal with staying sane. I have added a list of my book recommendations to my blog in the Bootstrapper's Bookshelf section.

  13. 1

    Thanks for taking questions, Arvid. Did you acquire customers for FeedbackPanda though a sales team or online channels or both? What contributed most to the buying decision?

    1. 2

      We had no sales team whatsoever over the course of the business. Danielle, CEO and Co-Founder of FeedbackPanda, was permanently active in the social communities our customers would be part of. She was on Facebook, on Instagram, coordinating our blogging and content efforts while being a genuine part of the online teacher community. She was present, helpful, engaging and encouraging. You can call it "embedded sales" if you want, but it was actually building a tribe that would sell the product to their peers for us. We added a referral system that was highly effective, as teachers are very happy to share things, particularly when they benefit from them.

      What contributed most to the buying decision was twofold: the product came highly recommended by their peers, as word-of-mouth was the prevalent form of marketing we had going on. In addition to that, our product was highly collaborative: teachers would share their feedback templates with each other. So a new customer would have access to hundreds of thousands of templates, pre-written, pre-validated by other teachers just like them. Once a customer understood that, it was an easy sale.

      We automated as much as possible. This was a low-touch business, almost no-touch for some customers. I would guess that over 50% of our customers never talked to our customer support over their journey using FeedbackPanda. Onboarding, conversion, payment, self-help support, all of this didn't require human intervention.

      1. 1

        Thanks so much for the detailed reply, Arvid!

    2. 2

      It was all word of mouth (taken from the very good IH Podcast with Arvid from Dec 2019). Product was for teachers, Co-Founder was a Teacher, she recommended it to her peers and off it went.

      1. 2

        Indeed! Our most effective marketing effort was mentioning FeedbackPanda in a comment to a Facebook post about who teachers solve their feedback problems. It snowballed from there. Teachers began sharing links to the product, and would reliably bring it up whenever feedback was discussed. So we didn't have to do much there. One post was enough to get it going.