Discover more from EcZachly Data Engineering Newsletter
How I quit my $600k data engineering job at Airbnb and made $600k in 7 months from content creation and teaching
In October 2022, a drunk-driving Tesla crashed into my house. This event pushed me to rage-quit San Francisco. I knew something was wrong in my life but I couldn’t quite figure it out.
I decided to become a digital nomad hoping my mental peace would improve. After visiting Denver, the Netherlands, Portugal, the United Kingdom, Portland, and Salt Lake City, I realized it wasn’t where I lived that was the problem; it was what I was doing.
Fine-tuning Airbnb’s marketplace by optimizing supply and demand wasn’t sparking the fire in my soul as it once did. I was intrigued by the notion of becoming a full-time creator and teaching the world about data engineering!
Getting inspired to make the shift
In January 2023, I needed inspiration for this new creator lifestyle. I lived with two YouTubers, Tina Huang and Ken Jee. After some jiu-jitsu, skiing, and many trips to the largest Costco in the world, they showed me clearly that the way I was living wasn’t the only way to be successful. It made me question my life path in the best possible way. This new-found wisdom gave me a strong sense of urgency to quit my $600k data engineering job at Airbnb!
Specifically what they showed me was:
Having complete ownership of your ideas is how you build long-term wealth
You have to diversify your income to find stability as an entrepreneur
Building community matters a lot
In this article, I’m going to talk about my journey over the last seven months in achieving my dreams. More specifically:
Why and how I quit my job
How I made money and continued building my brand
How did I market my boot camp and course
How I kept improving my boot camp
Summoning the courage to quit my job
I was unsure about quitting Airbnb mostly because:
EcZachly Data Engineering Newsletter is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.
I was making great money at Airbnb
Airbnb has a great work-life balance
I never had to go into the office at Airbnb
My team at Airbnb was amazing
The motivating factors for quitting Airbnb were:
I felt like I wasn’t doing my best job creating long-form valuable content
Whenever I tried to do long-form YouTube videos along with my job, my mental health suffered
I learned about Airbnb’s boomerang policy that would let me come back in some time if I decided entrepreneurship wasn’t for me
I wanted to do one thing great instead of two things with half my energy
February rolled around, and I told my manager that I wanted to quit. He’s somewhat surprised since I got an “exceeds expectations” on my performance review at Airbnb. The exceeds would’ve given me a lot more stock in the company had I stayed. I set my last day to be March 19th so I could honor my time at Airbnb and give them a smooth handoff of my work.
The first two weeks after I quit, I felt pretty unhinged. I went on Joe Reis’s podcast and talked about data engineering, ADHD, and entrepreneurship.
I ground out a ton of code for TechCreator, the platform I had been envisioning for a long time. This platform would allow people to create a searchable archive of all your social media content.
If you’re interested in trying out TechCreator, please use this contact form and I can help you get onboarded. I want to make this the best platform for cross-platform creators yet! TechCreator initially was going to be my “newsletter” competitor platform until I found Substack and realized the recommendations and network effect they have is just too powerful to compete with.
Aside from coding, I started grinding out short-form TikTok and Instagram videos. I aimed to create three videos a day every day for March and April. This laid the foundation for my short-form video success. I went from <2k followers on both platforms to over 15k in those two months.
The V1 boot camp
Near the end of March, I realized that TechCreator, TikTok, and Instagram weren’t going to make me much money and I started feeling the suffering of not making $50,000 every month. I decided to launch a data engineering boot camp to make some money. I posted about it and took on 50 students charging $987 each. I picked this price point because then I’d be making about the same as I was when I was working at Airbnb in a month.
I didn’t want my boot camp to be like other “breaking into data engineering” boot camps. I wanted to create a boot camp based on hard-earned experience. Experience lessons that you can’t just look up with GPT-4. I partnered with Sarah Floris on this boot camp. She did presentations on career skills.
Some of my favorite students from the V1 boot camp were Mitali, Julie, Scott, and Tony. Without Julie’s help, the boot camp would have been a disorganized mess! Lakshmi also helped a lot with homework help and grading in V1.
The V1 curricula covered data modeling, Apache Spark, Apache Flink, Tableau, and data quality. It was four hours a week of lecture/lab each week for six weeks. I loved teaching it. I had a ton of energy for the first two weeks but started to slow down after that.
Here’s what I learned:
Teaching Spark was difficult because I had never used Databricks before and I wanted to learn it. But getting 50 people set up with a Databricks trial account experience was way more annoying than I thought it would be.
I really enjoyed teaching fact and dimensional data modeling the most. Those weeks felt really magical and purposeful. I also taught about how I used data modeling to save Airbnb and Netflix many petabytes of storage in the cloud!
I needed a team to help me for the next boot camp. Running everything myself was too much
I hated sending invoices out to people and waiting for them to respond. Direct purchase seemed like a much better option
People loved the V1 boot camp! 84.6% of the respondents said it was worth the money and the rest said it was at least kind of worth it!
Seeing such a positive response from the V1 boot camp, I decided to set my sights much higher!
The V2 boot camp
I ended the V1 boot camp in early June and was tired. I decided to take a few weeks off to celebrate my wins. Around this time, the Databricks AI summit took off and I got really excited to build something out using GPT-4.
I set my sights on building out a checkout process with Stripe and NextJS using GPT-4 to collect payments for the V2 boot camp. I surprised myself and built everything out in two days. I launched the following day. We collected over $100,000 in payments between when I launched on June 30th and three days later on July 3rd!
I structured the pricing for V2 much differently from V1.
V2 was going to be twice the content of V1
V2 was going to have parallel tracks called Analytics and Infrastructure.
The price points were $1300 for one track and $1800 for both
I figured paying $500 more for twice the content would nudge more people toward the combined product and paying more.
I was right and over 80% of the boot camp attendees picked Combined.
I was much more thorough with marketing the V2 boot camp as well.
Stephanie’s shoutout on LinkedIn was incredibly powerful and we had another $30k day later in July.
I collected a great list of speakers to talk with in the V2 boot camp. This helped sales a lot too!
Stephanie Nuesi, Analyst at Google, content creator with 300k followers on LinkedIn and Instagram
Carly Taylor, Senior Machine Learning Manager at Activision with over 100k followers on LinkedIn
Sundar Velayutham, Senior Staff data Engineer at Apple worked at petabyte scale for many years
Parth Parekh, Data Engineering Lead, Generative AI at Meta. Worked with him for two years at Facebook and leveled up a TON!
Jitender Aswani, VP at StarTree. Jitender was my manager for four years at Netflix and Facebook. I saw my compensation grow from $170k to $550k in those four years!
Bill Inmon, Father of the data warehouse. Prolific author. Coined the term “data warehouse”
After the sales effort, we launched on July 16th with 122 signups! We had 10 people get refunded in the first two weeks and I felt bad because I thought I was building something great. But I realized that they were probably not ready for the boot camp and it wasn’t about what I was building!
I hired Julie, Lakshmi, and Sarah part-time again for this round of the boot camp!
Two weeks into the boot camp I bought dataengineer.io as a domain for $7000. It felt more fitting than eczachly.com.
I ended up teaching all twenty-two lectures in the V2 boot camp and filming sixty hours of content in six weeks!
Product development efforts during v2
We made tons of great improvements to dataengineer.io during the V2 boot camp:
I realized during V2 that I made the videos on dataengineer.io downloadable with one click. This was something I wanted to remove in order to protect my IP. I worked using FFmpeg and HLS encryption to transform my videos from .mp4 files to .m3u8 files so they couldn’t be easily downloadable.
We also made great improvements to the schedule tab and deprecated the use of Google Calendar. You can export a .ics file from the schedule tab now!
We added accountability from Zoom and certifications
We added time tracking to the videos so we could measure how much watch time was happening!
We added the mentor matching program to the V2 boot camp! For students that get certified, they get matched with a mentor who will help them succeed in their career free of charge!
The V2 boot camp officially ended on August 25th. 5 days before that, I made an Instagram Reel that had 10 million views! I grew from 30k followers to 110k followers in one week! The same video had 3.6 million views on TikTok and I went from 30k followers to 60k followers there!
This reel also caused a huge shift in my website traffic!
Launching my newsletter
In mid-June, I decided I needed to start writing more long-form content to connect with my audience on a deeper level! I imported all the emails I collected with the TechCreator product I had been building (~10,000 emails). This gave me a healthy base of emails to start writing my newsletter!
I wrote about:
Passing data engineering interviews
These articles helped my newsletter grow much faster and connect with people in their inbox in a more intimate way! Newsletter subscribers were 5-10 times more likely to buy my boot camp than any of my social media followers!
I knew one of the most important features of Substack was to get recommendations from other newsletters. Lenny taught me this from his crazy growth charts!
I partnered with many amazing newsletters like:
Software Engineering and System Design Newsletters
Slowing down to speed up
At the end of the boot camp, I immediately was off to Burning Man to take a much-needed break from working so hard! I ended up getting stuck in the mud there for a few extra days and it was such a great experience!
After I got back from the mud, September was dedicated to finishing up the pieces of the boot camp that were left over and teaching the LLM-driven data engineering bonus week. This is four videos you can watch all four hours of content here for free: day 1 lecture, day 1 lab, day 2 lecture, day 2 lab.
Sights set on V3 boot camp and product
At the end of September, I realized that I wanted to do one more iteration of the content before turning it into a subscription-based academy.
The big things I wanted to change between V2 and V3:
The data modeling weeks needed to run in the cloud, not on Postgres. A lot of the data modeling concepts I taught were only really relevant in the cloud and not really relevant in relational database land. I partnered with Starburst and Tabular to accomplish this without jumping head-first into DevOps hell!
I wanted to partner with big brands such as Databricks and/or Snowflake to sponsor the boot camp so we could reduce the cost on students and still make good money for the value we’re delivering. I wasn’t able to get any sponsorship for the v3 boot camp although future iterations are looking really promising!
I wanted fewer speakers but deeper conversations, instead of twelve speakers with 30 minutes each, we are doing six speakers with an hour each!
Joe Reis talks more in-depth about data modeling
Ben Rogojan talks more in-depth about becoming a data consultant and content creator
I wanted to open the boot camp up and make it international by offering purchasing power parity discounts for a select group of students. I did this for 40 students and made an additional $30k in the V3 sales!
I wanted to make the boot camp have significantly less setup so the subscription-based people could get immediate value.
I wanted much less manual onboarding for GitHub and Discord. We’re accomplishing this with Discord and GitHub automation.
Automation of group assignments based on timezone proximity and track. This is a crazy bin-packing problem that I solved with just SQL here.
When I cracked $600k in net sales, I felt accomplished! I took on all this risk in 2023 by quitting my stable, high-paying job but was able to make more money in about half the time.
A business is not a business without expenses though!
Over the last 7 months, I’ve incurred the following expenses:
These expenses come after my net income on Stripe:
I paid out ~$45,000 to employees.
I paid $7,000 for dataengineer.io
I paid ~$500/month for server costs so about $3,000
I paid $1,500 for large neon signs
I paid $1,200 for recording equipment
I paid $1,000 for fancy hoodies
I paid ~$2,500 for dinners, gas, hotels
I paid ~$1,000 for Zoom, Loom, JetBrains IDEs, Notion, etc
These expenses are already deducted from my net ($659,265 gross revenue):
I contribute 1% of my gross revenue to carbon capture for climate change (~$6500)
I gave Stripe ~3% of my income which accounted for ~$20,000
I gave Substack 10% of my newsletter income for ~$1,500
I gave back ~$30,000 in refunds
So, if you want to calculate some profit margins:
Total expenses were ~$90,200 (and $30k refunds)
$659,265 - $30k refunds - $90,200 = $539,065 net profit
$539,065 / ($659,265 - $30k refunds) = 85.6% profit
These calculations illustrate why I didn’t use a course hosting platform like Maven. Had I used a platform like that which takes 10%, I would have paid an additional $65,000! Maven would have cost me more than all my employee costs combined!
Kajabi almost won me over for hosting my course since it has a flat monthly fee but I figured having my own system would serve me better in the end since I’ll have complete freedom to do whatever I want!
The start of the V3 boot camp
We started V3 with our biggest live cohort yet, 133 students from 14 countries!
Going into V3 I was worried that Starburst and my web server weren’t going to handle the huge influx of queries. During the first week of classes, my students ran over 6300 Trino queries and it worked just fine!
My students this time around seem extremely engaged. Dozens of questions are being asked at the end of each session compared to four or five in the initial boot camps.
As I’m writing this newsletter, we are about to go into week 2, fact data modeling and I’m incredibly excited to see what is in store!
How this changed me
Initially after quitting my job, I was stressed out. I kept thinking, “Wow what have I done?” The uncertainty of the future felt very scary and I had so many doubts in myself. I felt like I wouldn’t be able to do it!
My resting heart rate soared in March and April after I quit because I needed a bit more certainty in my life. After V2 boot camp, I was able to find a groove that felt very comfortable and healthy. That is until the V3 launch.
Part of my body told me, “Don’t launch V3 this year Zach. You’ve done enough and you deserve a break.”
The ambitious side of me was like, “V3 is exactly what will put your content to the next level!”
Ambition won this time and my body protested. I ended up getting sick with strep throat and my resting heart rate soared again! After the launch and the first week of classes were over, I felt much better.
Beyond the physical, my mentality has shifted quite a bit. I feel like I’m finally entering the “entrepreneur” mindset instead of the employee mindset. This shift has helped me see money as an enabler and spending it is encouraged.
My appetite for risk has been increased. I realized that I can do things with uncertain outcomes because that uncertainty means a much higher upside if it works!
The future is very bright!
Thank you so much for reading this newsletter! I’m sorry there was such a gap between my last article and this one! I’m going to get back to a regular cadence so I can keep delivering you value in a way that is consistent and delicious!
If you’re interested in taking my sixty-hour self-paced course, please check it out here. You can use code 100TBS at checkout to get 10% off!
What should I have done differently in my journey to $600k? Should I have used more managed services? If so which ones? What do you want to see from me in 2024?
Make sure to share this story with your entrepreneur friends who are looking for some inspiration!
EcZachly Data Engineering Newsletter is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.