$100K, 31 days. Bootstrapping dreams coming true

The insane story of how my friends and I built one of the fastest growing AI startups in the world

You read the title correctly. This is not clickbait. I've been an Indie hacker, bootstrapper, founder, or whatever you want to call it for the past 4 years now, and I feel like I'm just beginning to cash in on the skills I've been developing along the way. This story is a pretty wild one, and I'm almost certain you're gonna learn a thing or two. I only see this as a great start to a long journey ahead filled with a bunch more highs, lows, laughs, and lessons and I'm gonna document all of it. If that seems interesting, and you're not subscribed be sure to do so!

Part 1: The Idea

It all started towards the end of December of 2022, I was in the process of deciding to sell my gamified meditation app and was thinking about what to do next. I have a backlog of startup ideas, and the one I kept thinking about wasn't all that special: an AI writing tool. I can hear the groans already, another AI writing tool? Are you dumb? Do you not see Jasper? Copy.ai?

Hear me out, it all started on Twitter where I began reading about a company that was crushing it in the B2C AI writing space, and was making tens of thousands in MRR, all through organic Tik Tok marketing. I checked out their product and saw a gap I thought I could take hold of. All of these AI writing tools make you use their web app, however, billions of people happily use Google Docs, Notion, etc. That's when I had the idea of making a chrome extension to augment the capabilities of AI into your favorite text editor.

I began doing more research and making quick prototypes on Figma. This was around the time ChatGPT was released, but I wasn't too worried due to the fact that ChatGPT required constant context switching, so I assumed my product's main selling point still remained. On top of this, we were going to position ourselves specifically for students.

I shared the idea with one of my close friends and one of the smartest people I know, Tejas, who I met in my last semester of college. He was totally on board with the idea as we were looking for an excuse to work together for a while. Tejas's roommate actually went to go study abroad, so I ended up moving in that same weekend. Weird feeling being on campus as a grad btw, but the speed of communication was well worth it.

We began hacking away immediately, and faced a lot of technical difficulties, but were able to get a prototype working in about 3 weeks. During this time we threw up a landing page and tried making some Tik Toks on ChatGPT, but weren't getting many views. This is where our third co-founder Dan comes in.

Part 2: Distribution / Almost Quitting

Me on the left, Dan in middle, and Tejas on the right

I also met Dan around the same time as Tejas, and he was also still on campus when I moved in. After grabbing a meal together, he suggested he could help us with creating content (has a Tik Tok account with about 700k followers by the way).

We happily accepted, the first couple of videos were experiments and did less than 5000 views combined. Around this time GPTZero was also released, and we are all devastated. If you're not familiar with GPTZero, pretty much a student from Princeton named Edward Tan developed a way to detect with high accuracy whether or not something was written by AI. We began discussing potential solutions, and even the possibility of exiting the AI writing space altogether.

There was one small beam of hope, but it seemed far off, what if we could rewrite the text until it bypassed Edward Tan's algorithm? It seemed kind of impossible how could 3 kids possibly defeat some of the brightest researchers in the world? We didn't have a choice so spent the night hacking away. After about 3 days we found a way to bypass it. Credit mostly to Tejas who read the research paper GPTZero was based upon and found the chink in its armor.

Part 3: the pivot

Before we even had an actual product, I began designing a prototype on Figma of our new AI Bypasser which looked kind of like a plagiarism detector, and pretty much rewrote your essay for you until it was undetectable by GPTZero. Using these prototypes, Dan began marketing them on Tik Tok.

We held our breath in anticipation, and to our pleasant surprise, the videos began blowing up like crazy. Tens of thousands beginning to join our waitlist. We quickly surpassed a million total views, and 30k sign-ups. The momentum would not stop, with each day being bigger than the last.

If you think about it though, the success kind of makes sense. GPTZero was brand new, and as a result, had millions of students in despair. We came onto the scene as pretty much the only solution, hopeful students began sharing our videos with each other. We switched from a red ocean to a blue ocean overnight.

Part 4: The Release

Tejas and I spent the next 7 days coding like madmen. Dan continued to push different Tik Toks / Instagram reels to increase momentum. The pressure to produce was extremely high at this time and we got into our fair share of arguments and fights. We released the bypasser alongside our chrome extension towards the end of February.

All the anxiety and doubts of the launch were there, would anyone really pay? What if our servers crashed? Was Stripe connected properly? A few minutes after launching we got our first paying user. We leaped for joy. Then our second. Then our third. On the first day, we made about $100 USD. We hadn't released an email campaign as we were still testing for bugs. Each day after day was bigger than the previous. We had close to 80k email contacts at this point. Unfortunately, our open rate for our launch email was only about 10%. Thankfully, it did not matter as much as people kept seeing us on their social feeds and checking the site out. In our first 5 days, we generated $1k and had about 5,000 users.

The next week was when it got wild. All our previously viral videos were getting pushed out by the algorithm again, which drove huge amounts of traffic to the point where our server was down for about an hour. Not a fun time. After our second week, we generated close to $50k in revenue and had close to 100k signups. This is where we began to see patterns such as huge dropouts on weekends in revenue and active users.

This past week alone we generated $50k. We were whooping for joy when we made our first $1000, but now we don't blink twice when we make $8000 in a single day. It's crazy to think how quickly the team acclimated to such high daily revenue numbers.

90% of our revenue is coming from the AI rewriter which took us a week to build, which is why even though we technically started more than a month ago on the chrome extension, the AI rewriter/bypasser from idea to $100k revenue took 31 days.

We’re grateful but definitely nowhere near satisfied due to a couple of big problems.

Part 5: The Problem / the Future

Right now we definitely don't have a sustainable business. Retention for our chrome extension isn't really there, our churn is relatively high and since most of our user base is students our revenue is going to plummet in the summer.

We're thinking of adding a B2B component to our business where we pretty much create custom models for businesses/journalists to help create AI that is more similar to their writing style/brand persona. We've talked to a few folks who wanted access to an API as well.

In addition to this, since we have about 200k emails of students, we're thinking of creating a gamified test prep app that pretty much acts as your personal tutor, with of course, a chatbot that grows alongside you. Either this or simply a ChatGPT IOS app, with something special I can’t really share yet.

We're not sure yet, all I know is this is just the beginning. We have enough funds to pretty much live anywhere for a couple of months during the summer, whether or not Tejas / Dan drop out of college is solely dependent on the next 5 months of our startup. We're applying to Y-Combinator and a couple of other accelerators, so fingers crossed.

Key Takeaways

• We could have spent months adding features to the chrome extension, when in reality what these students were dying for was a weapon against GPTZero.

• Timing is so so crucial, all three of us had the skills to execute when the opportunity became clear to us. We had years of experience building and marketing shit quickly. Luck = preparation meeting opportunity.

• Social media is literally free gold right now. Paid advertising is dying. Market your designs (smoke and mirrors) and see if an audience bites. Do not try to build something, and market it after. If you can figure out how to make content people want to watch, everything is so much easier. You go from pushing the boulder up the hill to letting it roll down.