I built a startup and filmed every minute on Youtube. Here’s what I learned.
Thumbnails from the YouTube series
After being part of a fast-growing YC startup, I want to show what the early days of making a product is actually like—so I’m building a startup product myself and filming the entire process.
Today, I’m launching the result—Shuffleboard, a new collaboration tool for real-time remote discussions. Think of it like a 10x simpler virtual whiteboard (or Jackbox for work meetings).
Shuffleboard is a solo SaaS project: I built it a day or two at a time over the last year and a half. Now the app is live, and real companies rely on Shuffleboard for running meetings, workshops, and retrospectives. This month I crossed a new milestone: people I don’t know have been converting to paying customers.
Before I started work on Shuffleboard, I knew I wanted to do something unique and challenging with this project. So every time I sat down at my computer, I recorded my face and screen, and uploaded the entire day of work YouTube. I’ve posted 40 of the 93 videos recorded so far (subscribe here to get updates).
Like in any startup, there’s a rollercoaster of ups and downs. The visual brand was designed in a couple hours, but figuring out how to secure the database took almost a week. There are moments where I’m having a blast, moments where I want to give up, and lots of hours where nothing much happens at all. But overall it’s been an overall incredibly rewarding experience.
Here’s what I learned.
#1. Start with a good idea
It’s kind of impossible to tell if a startup idea will work. Lots of the really successful ideas out there seem crazy at first, and lots of sensible-sounding projects go nowhere.
I picked an idea that seemed like it had a good chance of product-market-founder fit, or an idea that:
- I am uniquely qualified to work on — in my case, requires a lot of attention at the intersection between design and front-end engineering, my area of expertise)
- Solves a problem I already know a lot about — I run design workshops constantly as a consultant, and consistently feel the pain Shuffleboard solves
- Is niche enough that a big player probably won’t blow me out of the water right away
- Is targeted at customers that have money and a history of spending it — probably B2B, not a consumer product
- Is something that I could work on for years if successful
#2. Make space for what you want
There’s no way I would have been able to do this if I still had a full-time job.
Luckily I’ve been building a consulting business for the past three years, which allowed me to dial down my regular working hours and make room for projects like this.
(Also, I couldn’t have been successful in consulting without years of experience at startups, and it would have been hard to get that experience without the free time to teach myself to code, the support of my family, and a whole lot of other lucky things falling into place.)
#3. Book your hours in advance
I never accidentally worked on Shuffleboard. Every workday was blocked on my calendar at least one day in advance, and I did not accept any meetings before 4 p.m. on the days that worked.
When you are your own boss, you are also your own employee. You should have clear expectations about when you will work, and roughly stick to that schedule.
#4. Recording yourself can slow you down…
There’s a lot of extra work and mental energy involved in working in public. I had to keep my webcam background clean, mentally prepare myself each morning before getting on camera, record my face with my phone, record my screen, transfer and backup my massive video files, edit videos together, deal with corrupt files, upload massive videos, and track my work to write timestamps in video summaries. That’s all time that I could have spent working on the product instead.
#5. But it’s probably faster in the long run
The structure of having to perform live on camera, even when I was confident nobody will ever watch my videos, pushed me to make progress every day.
I was spending less time on my product — but the time I did spend was focused and rarely wasted. That means I spent fewer hours per week on code, but probably delivered more value per week than I would have.
#6. It forces you to finish tasks
“No zero days” was my mantra. Because I had to title each day’s video and start a new video the next day, I had to actually complete a meaningful task every day. That means not stopping halfway through, and trying to get each task all the way to ‘done’.
There were only a few times where a task took multiple days of work. None of my videos are called “maintenance” or “miscellaneous” or “cleanup”. I did have small tasks to do, but I only do them after my main job is done for the day.
#7. It helps you ship
We can all use some accountability pressure sometimes.
When you start sharing public videos of your project, the social pressure of failure forces you to show up and keep shipping. I don’t know if I would have shipped anything at all without the video series, and I probably wouldn’t be writing this blog post about launching.
My actual desk and recording setup
#8. Being on camera is so weird
I never thought I’d be someone to post my face or work online. I felt an absolute truckload of anxiety before recording my first video, and my body language shows it.
#9. Your energy will naturally ebb and flow
Most YouTubers can temporarily drum up some enthusiasm for a video shoot, but when you’re filming the entire process you can’t sustain that kind of energy. Many of my videos start out with “I’m tired and grumpy today… but whatever, let’s get started”.
The reality of work is that it’s sometimes exciting, sometimes boring and frustrating, and overwhelmingly emotionally neutral.
#10. Just be yourself, quirks and all
At first I assumed I had to take on an energetic persona for an audience, but I quickly realized it’s just not practical to “put on a face” for hours at a time and still be productive.
#11. Anything worth doing takes time
Even though I worked for almost 100 days, it took a year and a half to get those days in. That’s a huge investment of time, especially for one person.
A lot of smart people would say that’s way too long to work on something before launch. A lot of other smart people have gone the other direction, working full-time with multiple founders for years before launching.
I wish I had been able to work faster, but I also know this: easy-to-sell products are so freaking good that buying them is an obvious decision for customers. And really good products take time to make.
Maybe I’m wrong about this one. Maybe I’ll learn to be faster next time. But right now, I’m glad I stayed focused and patient.
#12. Iterate more than you invent
Everything I made was in a constant state of change. The more closely a feature was related to value for customers, the more times I had to redesign/rebuild it based on what I learned from users.
To show you how much work went into iterating, here’s a rough breakdown of how I spent time in my videos:
- Having an idea: 1 day
- Designing a crappy first mockup: 4 days
- Coding a crappy first version: 12 days
- Iterating & adding based on feedback: 78 days
#13. Build one piece at a time
Lots of people think software has to be done in one big sprint, but this is just wrapping modern jargon around the classic failed waterfall model.
Software is flexible when built correctly. You can make changes, adapt, and improve as you work, without knowing the finished product at the start.
Another way to say this is “paint in layers”: make broad strokes first, then worry about the fine details as you go.
#14. You can’t see the finish line from the starting line
Precise planning for the far future is an exercise for the ego. As a founder your vision should be about what you can do for customers, not what pixels go where. Your product will evolve as it interacts with the market, so don’t worry about predicting the future.
Starting a company isn’t like baking a recipe, it’s like going on a long trip. Don’t get too attached to the outcome in the early days, have a little faith, and it’ll be worthwhile.
#15. Talk to customers, early and often
Even though I didn’t do a press launch for 18 months, I was testing early versions with customers in the very first videos.
Launching and talking to users are very different things. Don’t wait until launch to talk to customers. By the time you do a public launch, you should already know that you’ve made something people love.
Talking to users isn’t too hard once you know how. Send some emails to a few people you know, ask them for nitpicks, then close your mouth and listen.
Shuffleboard is a tool for running simple real-time remote discussions and workshops
#16. Ignore the side comments
I’ve gotten a lot of comments from friends about my apartment’s ceilings, my long messy Covid hair, my plain t-shirts, my speaking mannerisms, my posting schedule, and more.
I think people just don’t know how to react when we see something different. We default to sharing our first-take observations out loud, even if they’re not the most constructive. These comments are pretty harmless. Just be ready to have the spotlight on you.
#17. Brush off the haters
The vast majority of YouTube comments I’ve gotten are short encouraging messages from random people I’ll never meet. (Thanks y’all ❤️)
But every once in a while, you’ll get a comment that comes across as pretty negative. Here’s someone on Hacker News implying that I wasn’t qualified to do this project.
My guess is that these commenters are still learning how best to communicate with other people. It’s not your job to unpack or solve this for them. As Brené Brown would say, they’re not “in the arena”, so their opinions don’t have to be interesting to you.
#18. Share even when it’s uncomfortable
When I say I put the entire process online, I mean the ENTIRE process. Theoretically, anyone could copy my whole app by watching my videos and typing what I type.
Will this have a negative consequence down the line? I don’t know! I hope not! All I know is that I made a bet that the benefits of being public would outweigh the costs. Right now, I’m glad I did.
If you’re working on a new project, peel back the curtain for people and work in public. I recommend using whatever format comes naturally. If you find yourself writing a lot, write about it. If you find yourself talking a lot, make a podcast or video. Be smart (don’t show your private keys, and consult your lawyer) but put your process out into the world.
#19. Some easy things are hard, some hard things are easy
I think we all have inaccurate preconceptions about how much effort new things are going to take to learn.
For example, before this project I had never written (or even read?) a press release. But writing one didn’t take long: I Googled around a bit, took a stab at a first draft, edited it after lunch, and got a friends’ feedback within about a day.
On the other hand, I spent far, far, far more time that I thought I would wrangling tiny bugs and getting my database set up in a way that worked for. (This is the reason that agile software proponents say that estimates are worthless.)
#20. Making is messy
One of the things I’m most proud of in the video series is that it shows the real-life, messy, distracted, sometimes even half-baked nature of the creative process.
Great ideas get thrown away, and mediocre ideas stick around longer than you might like. There’s no time, and yet everything takes far too long. Some things require compromise, others require a constant pursuit quality. You make a decision, you change your mind — then you change it back. You‘re questioning your decisions, reworking, massaging, and trying again and again and again — all while fighting perfectionism, laziness, fear, and boredom.
This is the inevitable side of making things that not everyone talks about. It’s also the part I’ve come to absolutely love. I know this is cheesy, but I love making things — it’s a wild wonderful weird adventure!
#21. Just do it
There will always be plenty of reasons not to take a risk: wasted time, potential embarrassment, waiting for the right moment, yadda yadda yadda.
At some point, you just have to just say “screw it” and do stuff anyway.
Remember, you don’t need anyone’s permission to make something. (Especially on the internet.)
Just start, and watch what happens.