My Experience Attempting to Launch a Startup

During my job search, I felt a transformation in my spirit and the outcomes of interviews after I shifted my time to building something. After a month of 100% focus on finding a new job, I decide to invest some of this energy in creating a product. Reframing this break from hiatus to opportunity, I saw it as fertile ground to cultivate new skills, undertake new challenges, and possibly build something innovative. In the spirit of a useful article, I will start with lessons I gained through this experience that are helpful for anyone currently job searching or thinking about building something themselves, followed by a detailed account of what I did.?

Throughout the article, when I reference what I built, I am talking about Quie, a research plan creation tool. It utilizes GPT to help product people build useful research questions and debiased interview questions. My aim for the tool was to empower all product people to conduct user research themselves, allowing dedicated user researchers to focus on the most complex research questions. While I have chosen not to pursue developing this into a VC-backed startup, I plan to continue refining the tool and likely offer it bundled with professional services down the road.?

Primary Learning Lessons

I had three primary learning lessons that made the experience worth it: Startups are just builders and sellers; Building a business is an emotional rollercoaster, and not every company should be VC-backed.?

Startups need Builders and Sellers: During this journey, one of the critical realizations was understanding the dual functions at the heart of any startup: building the product and selling it. These two roles are distinct yet symbiotic, equally critical for a company's success. As a solo founder, I played both roles which required tremendous growth. On the building side, I needed not just to engineer a solution but determine how exactly to build something as close to my design as possible. On the sales side, I am comfortable understanding user problems but not as much translating this to how much someone would be willing to pay (if anything). Moving forward, I will apply this understanding to make my research more impactful, recognizing that valuable insights must not only connect to one of these two primary functions but must be effectively communicated in this way to sell its implications to stakeholders.

Startups are Emotional Rollercoasters: Starting a company is an emotional journey that is as intense, if not more so, than a job search. Throughout my career, I've experienced the highs and lows of job searching: the thrill of the pursuit, the suspense of interviews, the exhilaration of an offer, and the discouragement of rejection. While building my generative AI solution, I discovered that starting a company pushes these emotions to an entirely new level. Each day, I woke up with an ambitious to-do list and the prospect of tangible daily accomplishments, which initially was very invigorating. However, with time, the reality of entrepreneurial challenges started to hit hard. There were incredible highs, such as my product's development breakthroughs and positive feedback from early testers. But there were also severe lows, like roadblocks in coding or lack of progress in certain areas. This emotional roller coaster taught me resilience and the importance of maintaining emotional equilibrium during the entrepreneurial journey.

Venture Capital (VC) Backing is Not the Only Option: When I finished my proof of concept, I felt the next step was to begin conversations with venture capital firms to see if I could raise a seed round. While the allure of VC can seem attractive, it inherently necessitates a particular business model. To be VC-backed, a startup must identify a large market with a significant unmet need and subsequently develop a solution that not only addresses this need but also outperforms other existing solutions substantially. This hyper-fixation of being the 'best' in a vast market can sometimes stifle innovation. Moreover, there are alternative approaches to creating and scaling a product that may align better with certain ideas or industries. For instance, some products may be more effective when bundled with services, often called a Software-with-a-Service (SwaS) model. This model provides a more comprehensive solution catering to nuanced or specialized needs, providing value beyond the standalone software. Such products might not yield the 'unicorn' status that VCs often hunt for, but they can still prove to be incredibly successful and rewarding in their own right.

Additional Technical Learnings

Beyond the key learnings I shared above, I grew my technical capabilities in engineering and finance during this period. A highlight of these learnings is shared below, which I believe is useful for anyone working in the product space.?

Much more to Generative AI than ChatGPT: The journey of building a generative AI solution propelled me into a rapid learning phase. From understanding the nuances of AI technology to implementing it in a real-world product, I immersed myself headlong into this fascinating world. I acquainted myself with various tools and platforms fundamental to the generative AI space, such as OpenAI API, Hugging Face, and LangChain. These platforms served as the bedrock of my technical operations, helping me create, manage, and optimize the AI components of my solution. This was key to my technical upskilling, broadening my toolset and sharpening my AI acumen.

No-code Platforms are Valuable: Building my solution also introduced me to the growing ease and accessibility of no-code platforms. These platforms are making it increasingly feasible for non-technical founders to bring their ideas to life, breaking down barriers and democratizing the creation process. I was astounded by how these platforms simplified complex operations, allowing me to focus more on the design and user experience. This was a significant revelation, highlighting how technology is making itself more accessible and fostering a more inclusive ecosystem.

Lots of Finance: On the financial front, it became evident that fundraising is a dominant theme in the narrative of most VC-backed startups. It was eye-opening to realize that founders invest a substantial portion of their time thinking and strategizing about fundraising. This added another dimension to my understanding of the startup ecosystem and highlighted how crucial it is to grasp the intricacies of fundraising and money management, irrespective of one's role in the organization. In a month, I learned how to conduct a Total Addressable Market analysis, predict Customer Acquisition Cost and Life-Time Value, and argue why a certain go-to-market strategy is preferred. These learnings have significant implications for anyone building a product to understand the underlying business case for why a product will be appreciated and purchased by potential customers… aka, even the most beautifully designed app will fail if it doesn’t address a real market need.?

My Experience Building and Pitching Quie

My adventure in building a product began the first time I used ChatGPT to write additional research questions I should consider when building a plan. While these systems work by regurgitating information and thus seldomly create novel insights, much research asks the same questions in different contexts and applications. Thus, at that moment, I realized the user researcher profession will undergo a significant transition as less skilled product people can conduct research more, and dedicated user research can become more efficient.?

A few months later, I put this theory to practice when completing a research plan as part of a freelance consulting engagement. My project involved completing a set of interviews for a company building a grant writing tool. While this is an area I have some familiarity with, I am far from an expert, and thus I used ChatGPT to train me on the basic processes and challenges. Taking this knowledge, I began writing research questions that ChatGPT then built out and provided multiple versions of interview questions I could ask. All at once, I realized this could be a product in its own right.?

I started my product development process by engaging in insightful conversations with people in my network about how they were currently creating research plans and ways they wish there were tools to assist. This step served as a form of market research, providing valuable insights into existing solutions, their limitations, and opportunities for improvement. Through these dialogues, I started identifying potential avenues to tackle the issue.

Zeroed in on the problem and potential solution, I built the first prototype. I had to expand beyond what ChatGPT could do alone so instead of building a simple chat interface, I scoped out a research question tracking software. Beginning with what I knew best, creating physical mockups that depicted the ideal flow of the solution. Once confident about the concept, I transitioned from these physical models to digital mockups using Figma, making the design more interactive, adjustable, and sharable. In this, I learned the power of design systems for rapidly creating beautiful solutions.?

With initial mockups ready, I initiated a series of user interviews. I contacted friends, former colleagues, and other startup founders at Founders University. These conversations served as my product's early validation and feedback rounds, offering invaluable perspectives that guided my iterative design process. Following the interview feedback, I built the solution using a low-code platform, Bubble (shout out to them for offering free credits to founders). This process was an exercise in working within certain constraints, but it also allowed me to see the advantages of such platforms. The constraints forced me to prioritize and focus on core functionalities, while the ease of use allowed for rapid prototyping and iteration. Building the primary functionality of the proof of concept, I integrated the OpenAI API into the application, a learning experience in and of itself. The best part of this process was that whenever I encountered a challenge, I would begin my troubleshooting by asking ChatGPT to debug my code. When that didn’t work, I would reach out to others working with large language models, a testament to the community that forms around builders.?

Armed with a proof of concept, I returned to the interview stage with something tangible to show. I conducted further interviews using the PoC to gather targeted feedback and gauge user reactions to the full product mockups. With a working PoC and positive feedback, I began the fundraising process. I pitched my product and vision to several investors and progressed quite far in discussions with a venture studio. We ultimately agreed that our visions were too far off to continue, but it was an exhilarating experience. This phase of the journey introduced me to the challenges and thrills of fundraising.

Concurrently, the day came to take the product public. I launched it on LinkedIn and various networks I am a part of, inviting feedback, engaging with potential users, and building a community around the product. I also set up analytics to track user engagement and gather data to guide further development. Within a week, it had over 100 users each that spent around 2 minutes in the app. While these numbers were initially impressive, I soon realized that few used the AI functionality. Had I continued building, I needed to create a more intuitive interaction to assist users along what is a new process for them.

As the dust settled post-launch, I found myself at a crossroads, needing to decide the future of this venture. It was a culmination of all the work I had put in, the feedback I had received, the potential I saw, and my personal aspirations and constraints.

Why I Decided Against Building Full-time

After two months of development, I was left with a proof of concept with initial usage, a pitch deck and talk track, and ambition to solve the problem I had spent significant time investigating. At the same time, I was running low on the personal runway, was already exhausted by the time and effort I had invested, and was concerned about the tradeoffs a founder’s life entails. When I was offered a job at a company I have followed for years doing behavioral science and AI research, I was forced to decide to go all in or treat this as a side project.?

I had to think through what I did and did not like; on the positive side, one aspect of this journey that I deeply appreciated was the opportunity to grow rapidly in diverse areas. As a founder, I had to develop my sales, marketing, and general business operations skills. The immersive nature of startup life compelled me to step outside my comfort zone, taking on roles and responsibilities that I wouldn't usually encounter in a traditional job. Another fulfilling aspect of this experience was the potential to build a tool to empower the democratization of research efforts, something I am deeply passionate about. This passion inspired and satisfied my work, making the highs higher and the challenges worthwhile. Although, as I went down the path of investment, I realized this mission might not be best achieved through a VC-backed startup.?

Recognizing that my true desire was encouraging the democratization of research, I began to see that consulting and thought leadership might instead be the preferred methodology. While most people saw the potential in my proof of concept, few actually used it in their work. I hypothesize that few product people see themselves as researchers and thus see a research question creation tool superfluous. Seeing that a successful product in this space requires strong tech and a process change, I recognized this was a high-risk option, more than I was willing to accept after two layoffs in a year. At the same time, market uncertainty added another layer of complexity to the challenge. Not only was there financial uncertainty, but the state of the generative AI market was also unpredictable. My product would use OpenAI’s API at its core; if access to that changed, it could quickly end any progress. I realized the personal investment would be all-consuming with all this risk and uncertainty. Ultimately the final straw was when I decided I was unwilling to make the required trade-offs in other areas of life.

So, in the end, I chose to continue building as a side project and instead pursue my greatest passion even though that meant giving up the Founder's ambitions (at least for now). I view this in no way as a failure but instead as a tremendous growth opportunity. Starting my new job, I recognize I will be much more effective with the research I conduct, ensuring it leads to products that work for users and increase revenue.?

I will conclude with a few shout-outs: First of all, to my wife, Alita Joyce , who was a sounding board throughout this entire process, plus she taught me so much about design. Next to Pratik Stephen , Sparsh Mehta , and Alex Henderson , who were great advisors throughout this process, I only went as far as I did with the investors, thanks to this crew. During this period, I was also enrolled in Founders University, led by Kelly Schricker and Presh Dineshkumar , which proved to be a tremendous educational program and built further confidence that I could build something big. Beyond these specific callouts, thank you to all those willing to speak with me, provide feedback, or engage in my polls and posts, which were helpful in refining the idea.?