#11 Interviews - Alex Rainey
AI Bootstrapped Startup Generating $20K+ MRR
I had a chance to meet with Alex through his AMA on Reddit, here is where you can get more details.
After his experience in corporate life, he sold 4 startups and raised $1.5M. Now focusing on customer support AI tool with his co-founder and generating $20K+ MRR.
Can you give us some information about your background?
Certainly, my bootstrapping journey started after spending about five or six years working in startups, with my initial experience at Accenture in the digital department. During my time there, I acquired skills in technology, honed analytical abilities, and gained exposure to various aspects of building and managing teams. Despite the valuable experience, I began feeling a strong desire to embark on an entrepreneurial journey.
The first startup I co-founded focused on the insurance space, specifically travel insurance. Over two years, we raised approximately 1.5 million dollars in funding. However, during the pandemic, we faced challenges and decided to pivot from travel insurance to travel planning. This shift involved embracing a B2C approach with a mobile app. While the experience provided significant insights, the intricacies of the B2C space, such as customer acquisition and retention, posed substantial difficulties.
Facing funding challenges in 2022 amid the post-pandemic landscape, we made the strategic decision to wind up the business. Fortunately, we found a buyer for the app, ensuring its continued existence. Following this, I explored consulting opportunities but maintained a strong inclination to work for myself. During this period, I connected with my current co-founder, Mike Heap. Our discussions led us to explore ideas, coinciding with the release of GPT-3 by OpenAI.
This technological advancement inspired us to develop a university application writing service called the "ucast statement generator." While the ethical implications of using AI for application writing raised concerns, the service effectively showcased the capabilities of the technology. However, we soon found ourselves growing disinterested in the B2C focus, prompting us to sell the service on Acquire.com. This transition marked the beginning of our shift to a new project.
After selling the ucast statement generator on Acquire, what was the next step for your business?
Yes, that's correct. We eventually sold the project through Acquire, although it happened later after listing it. Following that, we embarked on a significant development in our current business journey. We introduced a product that enables users to fine-tune GPT-3 models with their own data by uploading a CSV file of content. At that time, fine-tuning was a challenging process, involving complex steps like converting data into a specific format and utilizing command-line interfaces.
The product addressed the difficulties associated with fine-tuning and showcased remarkable results. We had previously realized the potential of fine-tuning through the university statement writer, where it significantly enhanced the capabilities of the models, especially when dealing with queries that the standard model might not understand. The business gained momentum, and we officially launched it on Ben's Bites during a hackathon he organized in late 2022.
As we delved into the AI community, building connections and gaining visibility, the product garnered substantial demand and traction, leading to the generation of decent revenue. Subsequently, when the embeddings API was introduced, we were once again excited about the enhanced possibilities it offered. This API provided a powerful semantic search capability, making it easier to query documentation and retrieve highly relevant information from PDFs, websites, and more. The embeddings API essentially supercharged semantic lookup technology, opening up new avenues for innovative applications.
Sponsored Ad
For SaaS companies, one of the biggest challenges is onboarding and Inturact provides a complete and personalized user onboarding service for high-growth companies.
They have a special price for SaaStrappers readers and you can get 15% OFF when you mention our name in the email.
Here is the email address of Joshua — > [email protected]
At that time, were both of you focusing on these projects full-time?
Yes, indeed. We were both fully dedicated to these projects. We had some savings from consulting work, and we started earning additional income through small, one-off development projects for clients. When certain businesses were sold, we received lump sums of cash, ranging from ten to twenty thousand dollars, which helped sustain us until we reached a break-even point with our current business.
Were you primarily using your savings during that period?
Pretty much, let’s say 50-50. It was a mix of using our savings and generating income from client projects. It was a balanced approach, with both aspects contributing to our financial support.
Afterward, did you shift your focus to the new AI project?
Yes, we initiated a pre-sale and early access pre-sell for the fine-tuning product. We offered users the ability to chat with their data, enabling them to upload content and ask questions. Surprisingly, what users were truly interested in was creating a customer service chat or an intelligent assistant that knew about their company. This insight led to a pivot, and we garnered around 10,000 pre-sales, demonstrating exceptional proof of demand. This prompted us to halt everything else and concentrate on building this new product. We launched the current product in early March, a rudimentary chat-with-your-data application, and since then, the journey has been nonstop. While we aim for faster growth, we've achieved significant revenue and customer milestones. Currently, it's just my co-founder and me, along with some freelancers for development. We utilize a no-code platform for most of our product development and are committed to moving forward swiftly. We're enjoying the journey and have no immediate plans to sell or seek investment.
Have you been using Bubble for this project?
Yes, we utilized Bubble as our platform of choice.
After March, how did you approach marketing for this product?
The most significant strategy, unintentionally, was tapping into the customer base of our fine-tuning product, which had around 15,000 users. We had a mailing list with those users, and porting over those interested in our next venture resulted in around 10,000 pre-launch sales. This provided our initial boost, leading to our first few hundred customers. Beyond that, we engaged in regular activities like being active on Twitter and LinkedIn. While our Twitter audience was smaller, we were gradually growing it. LinkedIn, with its larger audience from our professional services backgrounds, proved to be a key platform. We also attempted engagement on Reddit, faced challenges, and explored AI directories and newsletters to reach the early adopter crowd.
Subscribe to our newsletter:
To receive weekly interviews,
To get the latest updates about SaaS Bootstrappers.
Are you currently fully focused on this project, and what are your future plans? Are you considering continuing with bootstrapping or seeking VC funding, given your past experience with VC funding?
Yes, we are fully focused on this project, and the current bootstrapping model is preferable to our past VC-funded experience. In the previous startup, constant reliance on investor money created a cycle of fundraising, stressing about investor updates, and managing investor relationships. Now, with sufficient revenue to cover costs and draw salaries, we aim to remain bootstrapped. While this path isn't always feasible, we want to stay lean, keep the team small, and automate processes. Our goal is to continue bootstrapping and avoid raising external investment.
How does your current team structure look?
Our team consists of just me and Mike, the co-founder. Additionally, we have two developers who work as freelancers, dedicating a few hours per week to the project. This lean structure allows us to maintain flexibility, keep costs low, and focus on efficient operations.
In your AI startup, you mentioned that the churn is higher compared to B2B SaaS and falls between 10-20% in one of your Reddit posts. Could you elaborate on how you handle this churn, and do you believe that a misalignment between user expectations and the actual capabilities of the product contributes significantly to this churn?
Absolutely, the churn rate in our AI startup hovers around 12-13%, which is notably higher than the typical churn seen in B2B SaaS (usually ranging between 1-3%). It's a challenge we actively address, and a substantial portion of this churn is attributed to what we label as a fundamental misalignment.
This misalignment often stems from users entering our platform with expectations that mirror the capabilities of well-known AI models like ChatGPT. However, our product serves a different purpose, and when users discover this difference, it can lead to disappointment and, subsequently, cancellations.
To proactively manage churn, we've implemented a well-structured cancellation flow. This process not only makes cancellations straightforward for users but also provides us with valuable insights into their reasons for leaving. We believe in making it easy for users to share their feedback and reasons for cancellation.
Furthermore, my co-founder dedicates a significant amount of time each week to engaging with both prospective and existing customers. This proactive communication strategy helps us not only to address concerns but also to gauge whether our product aligns with the specific needs and expectations of users. By staying close to our user base, we can better understand their issues, iterate on our product, and enhance the overall user experience.
In essence, while the higher churn in our AI startup presents a challenge, we are actively working on building a more aligned and satisfied user base through transparent communication, feedback mechanisms, and a commitment to addressing user needs.
As someone experienced in bootstrapping an AI startup, what advice would you give to fellow bootstrappers in the AI space, especially considering the narrative that emerging technologies from major players may shrink opportunities for startups?
Certainly, one prevalent trend I've observed is the narrative that major players, like OpenAI, are releasing new technologies that might be perceived as killing opportunities for startups. My advice is not to buy into this mindset. It's a perspective that, in my opinion, is flawed and often comes from individuals who may lack experience as founders.
The reality is, we're just scratching the surface of what platform providers like OpenAI can offer. The continuous release of groundbreaking features and services for both end consumers and developers will likely persist for many years. Providers like Microsoft and Google will also integrate AI into every facet of their applications.
If this evolving landscape discourages you, it might be challenging to navigate the AI space in the coming years. However, if you can set aside these concerns and focus on building for a specific niche of users, tailoring your solution to a distinct subset of problems, there will always be a market for your unique offering.
Here's the key: stay niche and highly focused. Don't aim to solve a broad range of problems or cater to multiple segments. As an indie hacker, your advantage lies in staying small, niche, and off the radar of larger companies. By remaining laser-focused, you can address a specific segment's needs in a way that big providers won't.
This approach allows you to build a business that might generate substantial monthly recurring revenue (MRR) in the tens of thousands, potentially leading to a life-changing exit. While building a unicorn may require a different strategy, achieving significant success in a small, specific niche is very much possible. The key is to hunker down, stay focused on your chosen segment and problem set, and resist getting distracted by the noise and disruptions in the broader AI landscape.
How do you approach differentiating your product, especially considering the challenge of maintaining a competitive edge in a rapidly evolving space like AI? Also, how crucial is it to build a significant "moat" for your business?
In the dynamic landscape of AI startups, building a substantial moat at an early stage can be challenging. I've been skeptical about the idea of having a significant moat at the outset, especially for technology SaaS products developed in 2023. The reality is that any early-stage business, unless it has fundamentally groundbreaking technology or a novel approach (e.g., in the medical space), is unlikely to establish a significant moat within a short period.
Understanding this, the focus shifts to what can truly differentiate the product. One of the primary differentiators is the founding team—the way they think, operate, and their ability to execute. Recognizing that any technology product can be replicated quickly, the key is to leverage the founding team's unique qualities and execution capabilities.
While competitors may replicate features, sustaining this strategy is not viable in the long run. The founding team's ability to stay close to customers, execute at an impressive pace, and consistently deliver value is what sets them apart. The emphasis should be on outperforming competitors, being better founders, and executing more effectively.
Building a moat, in this context, is less about specific features and more about the founding team's strength, focus on a specific segment, and their commitment to solving a particular problem. The sustainable advantage lies in the team's ability to execute, adapt, and innovate—attributes that are difficult for competitors to replicate.
In essence, rather than obsessing over the concept of a moat, strong founding teams should focus on execution, continuous improvement, and staying true to their vision. The key is not to be overly concerned about competitors replicating features but to consistently outperform and deliver unique value to customers.
Previous Issues:
Thanks again for reading the interview and don’t forget to share this issue if you want your friends to read.
Reply